Open Hub
Open Hub

 Settings |  Report Duplicate

0
I Use This!
Very Low Activity

Commits : Listings

Analyzed about 7 hours ago. based on code collected 1 day ago.
Nov 05, 2025 — Nov 05, 2026
Commit Message Contributor Files Modified Lines Added Lines Removed Code Location Date
1218. (THEORY and FEATURE) Music Synthesizer - Python 3.8 upgrade and Fractal implementation - 21 February 2022 - related to 3,298,642,643,644,659,681,1124 and all sections on Intrinsic Merit of Music, MIR(Music Information Retrieval), Music Similarity and Clustering, AI Music Synthesis, fMRI connectomes - OpenNeuro - Psychology of music, Weighted automata, Chaos, Deep Learning, Timeseries analysis, Physique recognition and Motion capturing algorithms, SETH
Krishna iResearch - NeuronRain - K.Srinivasan
More... over 4 years ago
1218. (THEORY and FEATURE) Music Synthesizer - Python 3.8 upgrade and Fractal implementation - 21 February 2022 - related to 642,643,644,659,1124 and all sections on Intrinsic Merit of Music, MIR(Music Information Retrieval), Music Similarity and Clustering, AI Music Synthesis, fMRI connectomes - OpenNeuro - Psychology of music, Weighted automata, Chaos, Deep Learning
Krishna iResearch - NeuronRain - K.Srinivasan
More... over 4 years ago
---------------------------------------------------------------------------------------------------------------------------------------------- 1218. (THEORY and FEATURE) Music Synthesizer - Python 3.8 upgrade and Fractal implementation - 21 February 2022 - related to 642,643,644,659,1124 and all sections on Intrinsic Merit of Music, MIR(Music Information Retrieval), Music Similarity and Clustering, AI Music Synthesis, fMRI connectomes - OpenNeuro - Psychology of music, Weighted automata, Chaos, Deep Learning --------------------------------------------------------------------------------------------------------------------------------------------- 1218.1 Music Synthesizer in python-src/music_pattern_mining/AudioToBitMatrix.py (from Functions,Automata and Randomness) has been enhanced to generate music notes waveform from a fractal dataset. It has also been upgraded to Python 3.8 by 2to3 and autopep8. 1218.2 Function notes_to_audio() has been updated to include additional boolean argument for enabling fractal waveform and a clause for computing the chaotic Mandelbrot set (Real only) from a logistic function. Samplerate and Amplitude for the waveform have been set accordingly for loading .wav file in Audacity. Fractal set of notes is grown within a loop of eval() by feedback. 1218.3 Fractal music waveform is written to separate .wav file - function_synthesized_music_fractal.wav. Scaled notes are cast to NumPy int16. 1218.4 Amplitude is multiplied by array of notes by NumPy Array broadcast throughout for all types of music synthesis - function,fractal,automata and randomness. Music waveform has been sufficiently amplified now and audible in Audacity. Waveform could be loaded in Audacity and analyzed visually by zooming the track. 1218.5 Three invocations of music synthesizer below respectively generate music from a quadratic equation, sinusoid and a Fractal Verhulste-Pearl-Reed(VPR) logistic function : notes_to_audio(function='(x*x+x+1) % 32767',fractal=False) notes_to_audio(function='int(math.sin(x*x+x+1) * 32767)',fractal=False) notes_to_audio(function='5*(x*x-x) % 32767') 1218.6 Following are logs and music waveforms synthesized from finite state automaton,function,randomness and fractals: automaton_synthesized_music.wav function_synthesized_music.wav notes_synthesized_music.wav function_synthesized_music_fractal.wav testlogs/AudioToBitMatrix.log.MusicSynthesizer.21February2022
Krishna iResearch - NeuronRain - K.Srinivasan
More... over 4 years ago
+--------------------------------------------------------------------------------------------------------------------------------------------- +1217. (THEORY and FEATURE) Fictitious Neuro Cryptocurrency Mining Rig - Random Integer Partition + Rectangular Area Factorization + Money Changing Problem Proof of Work - Digital Watermarking implementation for Visual Neuro cryptocurrency - 15 February 2022 - related to 112,113,114,488,751,784,785,787,791,841,848,859,1095(a,b,c,d),1215,1216 and all sections on Economic merit, Set Partition analytics, Pseudorandom generators, Computational Geometric Factorization, Money Changing Problem - Frobenius Coin Diophantines, Computational Theory of Value, Poverty Alleviation, Neuro Cryptocurrency implementations in AstroInfer and KingCobra, Optimal Denominations,Computational Economics - Computational Hardness (Labour) Theory of Value, Histogram and Set Partition analytics, Equidepth-Equiheight Histograms, Separate Chaining Hashtables, Timeout gadgets (Refcounted Expirable C++ objects, KingCobra atomic refcounting and Survival Index Timeout) in NeuronRain, Digital Watermarking, Drone Electronic Voting Machines, Numeric Compression by Prime power encoding - Unique Factorization Domain, Multiplicative Partitions, Compressed Sensing (Succinct representation), Large Scale Visuals, Cybercrime analytics +---------------------------------------------------------------------------------------------------------------------------------------------
Krishna iResearch - NeuronRain - K.Srinivasan
More... over 4 years ago
--------------------------------------------------------------------------------------------------------------------------------------------- 1217. (THEORY and FEATURE) Fictitious Neuro Cryptocurrency Mining Rig - Random Integer Partition + Rectangular Area Factorization + Money Changing Problem Proof of Work - Digital Watermarking implementation for Visual Neuro cryptocurrency - 15 February 2022 - related to 112,113,114,488,751,784,785,787,791,841,848,859,1095(a,b,c,d),1215,1216 and all sections on Economic merit, Set Partition analytics, Pseudorandom generators, Computational Geometric Factorization, Money Changing Problem - Frobenius Coin Diophantines, Computational Theory of Value, Poverty Alleviation, Neuro Cryptocurrency implementations in AstroInfer and KingCobra, Optimal Denominations,Computational Economics - Computational Hardness (Labour) Theory of Value, Histogram and Set Partition analytics, Equidepth-Equiheight Histograms, Separate Chaining Hashtables, Timeout gadgets (Refcounted Expirable C++ objects, KingCobra atomic refcounting and Survival Index Timeout) in NeuronRain, Digital Watermarking, Drone Electronic Voting Machines, Numeric Compression by Prime power encoding - Unique Factorization Domain, Multiplicative Partitions, Compressed Sensing (Succinct representation), Large Scale Visuals, Cybercrime analytics --------------------------------------------------------------------------------------------------------------------------------------------- 1217.1 Neuro Cryptocurrency Mining Rig in Streaming_SetPartitionAnalytics.py has been augmented to print a visual Neuro currency which is digitally signed by Digital Watermarking OpenCV2 overlay of Neuro currency symbol (single strikethrough-N) and name. 1217.2 Digital Watermark is a configurable image in testlogs/NeuroCurrencyWatermark.jpg (contains Neuro name and symbol). Blank cleanslate currency template is stored in testlogs/NeuroCurrencyTemplate.jpg. Neuro cryptocurrency mint proof of work, a combination of Random Integer Partition + Lagrange Sum-of-squares Partition-to-Rectangular Area Reduction and Factorization + Money Changing Problem (BPP pseudorandom choice of an integer partition from SymPy, MRC or NC factorization of Lagrange Sum-of-squares Integer Partition-to-Rectangular Area Reduction, NP-Complete solutions to two Money Changing Problem Frobenius Coin Diophantine Integer Linear Programs equated to factor sides of the rectangle - and thus effectively a non-trivial pipeline of BPP + MRC-NC + NP) computes the factor sides of the rectangle equal to value of Neuro cryptocurrency and solves two ILPs. 1217.3 All factors of the integer factorized (= value of Neuro cryptocurrency mined) are concatenated and encrypted to SHA256 cryptographic hash as UUID of the currency - Currency value and SHA256 UUID are written to testlogs/NeuroCryptoCurrency.jpg based on template testlogs/NeuroCurrencyTemplate.jpg by OpenCV2 putText(). 2 leading "ff"s requirement as non-trivial Proof of Work in a pseudorandomly chosen Boost UUID is thus relaxed theoretically by uniqueness of integer factorization into multiplicative partition of prime powers (Factorisatio Numerorum) which has been already implemented in NeuronRain Grafit in the context of compressed succinct storage of a huge integer. Euclidean Semiprimes could be an apt choice as currency values as they could be factored into 2 primes and solved for 2 Money Changing ILPs. 1217.4 Visual Neuro Currency object testlogs/NeuroCryptoCurrency.jpg thus dynamically printed is watermarked by testlogs/NeuroCurrencyWatermark.jpg and written to testlogs/NeuroCryptoCurrency.jpg_Watermarked.jpg by watermark_image() function of DigitalWatermarking.py implementation in NeuronRain Grafit. An example Neuro Visual Currency Note of value 100 and SHA256 UUID watermarked by Neuro symbol and name backdrop is unveiled at testlogs/NeuroCryptoCurrency.jpg_Watermarked.jpg. 1217.5 Proof of work computation earlier has been captured in testlogs/Streaming_SetPartitionAnalytics.NeuroCryptocurrencyWatermarking.log.15February2022. Some commented code on EVM analytics has been uncommented to regenerate JSON persisted Drone EVMs - testlogs/Streaming_SetPartitionAnalytics.EVMs.json 1217.6 Expirable C++ Objects which could be timed out or expired by reference counting have been implemented in NeuronRain AstroInfer cpp-src/expirable/ and described theoretically in Conceptual Survival Index Timeout WCET OS Scheduler and KingCobra Message-As-Currency Atomic transactional Refcounting (over cloud) within KingCobra linux kernel driver. Refcounting is a bidirectional logical clocktick for objects oscillated by increment and decrement - objects are garbage collected by reaper when refcount is 0. C++ expirable objects which were meant originally for limiting number of views of an Image object (e.g a photo sent over email self-destructs after constant number of views,a privacy feature most sought after),could be suitably used for Visual Cryptocurrency sent over networks as add-on security to limit the transactions involving a currency ID. 1217.7 Neuro Cryptocurrency mining rigs theoretically demonstrated earlier try to equate value of a currency in terms of Hardness of computation and simulated value of a commodity which is producible by similar real-life hardness and are potential poverty eradication tools.
Krishna iResearch - NeuronRain - K.Srinivasan
More... over 4 years ago
Section 1206 updated
Krishna iResearch - NeuronRain - K.Srinivasan
More... over 4 years ago
Section 1215 updated
Krishna iResearch - NeuronRain - K.Srinivasan
More... over 4 years ago
1215. (THEORY and FEATURE) Set Partition Analytics - Set Partition-Rectangular Area to Factorization reduction and Integer Linear Program sol ution of Frobenius Coin Diophantines for Neuro Cryptocurrency mining rig Proof of Work, One Time Password facility augmented for Voter Receive d Encrypted Paper Audit Trail (VREPAT) Conceptual Drone Electronic Voting Machine - related to 135,1140 and all sections on Money Changing Pro blem,Optimal Denominations,Integer Partitions, Computational Economics - Computational Hardness (Labour) Theory of Value, Histogram and Set Partition analytics, Equidepth-Equiheight Histograms, Separate Chaining Hashtables, Drone EVMs, Space Filling and Teapot Shards, Equal and Unequal Circle packing, Grid filling, Integer Linear Programming and Cellular automata, Quantum Versus Classical Computation, Parallel Pseudorandom Generators, Berry-Esseen Central Limit Theorem - 11 February 2022
Krishna iResearch - NeuronRain - K.Srinivasan
More... over 4 years ago
-------------------------------------------------------------------------------------------------------------------------------------------- 1215. (THEORY and FEATURE) Set Partition Analytics - Set Partition-Rectangular Area to Factorization reduction and Integer Linear Program solution of Frobenius Coin Diophantines for Neuro Cryptocurrency mining rig Proof of Work, One Time Password facility augmented for Voter Received Encrypted Paper Audit Trail (VREPAT) Conceptual Drone Electronic Voting Machine - related to 135,1140 and all sections on Money Changing Problem,Optimal Denominations,Integer Partitions, Computational Economics - Computational Hardness (Labour) Theory of Value, Histogram and Set Partition analytics, Equidepth-Equiheight Histograms, Separate Chaining Hashtables, Drone EVMs, Space Filling and Teapot Shards, Equal and Unequal Circle packing, Grid filling, Integer Linear Programming and Cellular automata, Quantum Versus Classical Computation, Parallel Pseudorandom Generators, Berry-Esseen Central Limit Theorem - 11 February 2022 -------------------------------------------------------------------------------------------------------------------------------------------- 1215.1 Streaming_SetPartitionAnalytics.py has been updated to change following functions to expand their scope to Neuro Cryptocurrency Proof of Work (NeuroFed) and Drone EVM one time password padding facility respectively: def setpartition_to_tilecover(histogram_partition=None, number_to_factorize=1, solution="ILP",Neuro_Crypto_ProofOfWork=False) def electronic_voting_machine(Voting_Machine_dict, unique_id, voted_for, Streaming_Analytics_Bertrand=False, onetimepassword=None) 1215.2 setpartition_to_tilecover() now accepts an additional boolean flag Neuro_Crypto_ProofOfWork to treat the number_to_factorize argument as a Neuro Currency of same value and factorize the Currency value N=pq to obtain two Frobenius Money Changing Coin Diophantines which are equated to factor pairs p and q. 1215.3 Integer Linear Program CVXOPT GLPK solutions for the two diophantines are printed to partition the Neuro currency as sum of multiples of denominations. Computational Theory of Value (a rephrasal of Labour Theory of Value) equates Hardness of the supply chain producing 1 unit of commodity from raw materials to market and its commercial value. 1215.4 Agriculture usecase supply chain following Samuleson-Stolper convention : Seeds - Land rental - Farm labour wages - Water - Materials - End product. Computationally, agriculture is a Space filling problem of randomly sowing seeds in a 2 dimensional plane and their probability of fructification (a la Teapot Shards model) - Seeds parallelly sprout and grow ordained by a probability distribution of ordinates from nature's parallel pseudorandom generator (invisible hand of providence) and thus could be in NC or Randomized NC (Tygar-Reif Parallel PRG, Applebaum PRG in NC1 with linear stretch could churn out ordinates for seeds in parallel - NC - and only some randomly chosen seeds sprout in parallel - RNC). Finding total number of successfully sprouted seeds is tantamount to sum of random variables for the seed ordinates which tends to mean by Central Limit Theorem. 1215.5 electronic_voting_machine() now accepts an argument for One Time Password presumably a serial number of paper currency or Neuro Cryptocurrency Boost 512 bit UUID which is appended optionally to ripemd160 digest of Voter's public id context string. Defeating this uniqueness salting of an anonymous vote at a Drone polling terminal is as hard as counterfeiting of currency - paper or digital. 1215.6 Passlib SHA256 encrypt() has been deprecated and replaced by hash() while appending to a candidate's votes bin. 1215.7 Clauses have been introduced to differentiate Set partition to square tile cover reduction and Neuro Crytocurrency proof of work.An example Neuro currency of #8375 has been mined by Factorizing 8375 (Neuro currency has been assigned a logo "#" to distinguish it from legal tenders). SymPy Random integer partition of the Neuro currency value is reduced to a rectangle and then factorized. Logs testlogs/Streaming_SetPartitionAnalytics.Python3.9Upgrade.NeuroProofOfWorkILP.log.11February2022 print the minted Neuro Cryptocurrency and its optimal denominations as sum of products.
Krishna iResearch - NeuronRain - K.Srinivasan
More... over 4 years ago
Section 1214 updated
Krishna iResearch - NeuronRain - K.Srinivasan
More... over 4 years ago
1214. (THEORY and FEATURE) Computational Geometric Integer Factorization - 20523 bits quadcore benchmarks - Factorization as a non-trivial Pr oof of Work computation for Neuro Cryptocurrency mining rig - related to 620,713,728,739,1158,1159,1161,1213 and all sections on Neuro Cryptoc urrency, Computational Geometric Factorization, Set Partitions to Exact Square Tile Cover Reduction by Lagrange Sum of 4 Squares - Optimal den ominations and Money Changing Problem, Integer Partitions, Equidepth-Equiheight V-Optimal Histograms and Set partition analytics, Drone Electr onic Voting Machines - 9 February 2022
Krishna iResearch - NeuronRain - K.Srinivasan
More... over 4 years ago
-------------------------------------------------------------------------------------------------------------------------------------------- 1214. (THEORY and FEATURE) Computational Geometric Integer Factorization - 20523 bits quadcore benchmarks - Factorization as a non-trivial Proof of Work computation for Neuro Cryptocurrency mining rig - related to 620,713,728,739,1158,1159,1161,1213 and all sections on Neuro Cryptocurrency, Computational Geometric Factorization, Set Partitions to Exact Square Tile Cover Reduction by Lagrange Sum of 4 Squares - Optimal denominations and Money Changing Problem, Integer Partitions, Equidepth-Equiheight V-Optimal Histograms and Set partition analytics, Drone Electronic Voting Machines - 9 February 2022 -------------------------------------------------------------------------------------------------------------------------------------------- 1214.1 Non-trivial proof of work computation requirement for mining Neuro cryptocurrency could be satisfied by hardness of factoring large integers on MapReduce-BSP-Multicore-PRAM platforms - mostly semi primes and integers of low prime factors (Little Omega and Big Omega) - which are Nick's class and MRC formalisms and not NP-complete. 1214.2 Integer factorizations have been demonstrated earlier as side-effects of Equidepth-Equiheight Histograms, Set partitions to Lagrange Sum of Squares Tile cover reduction and Money Changing Problem for Optimal Denominations (Frobenius Diophantine Coin Problem) - Denominations which are decimal multiples of 1,2,5 and importance of 18 cent piece - any integral Neuro currency value can be expressed as integer partition or linear combination of multiples of denominations 1,2,5: Example: 18 = 1*1 + 1*2 + 3*5 1214.3 Set partition to Lagrangian Sum of Squares Tile Cover reduction results in two linear diophantine equations for each side of the rectangle (of area N=pq) which are equated to factorized pair p and q (each factor is a sum of sides of squares covering the rectangle) thus solving Money Changing Problem - Section 1161 exactly solves the two linear integer program Money Changing Problem diophantines (NP-Complete problem) by CVXOPT GLPK and Section 728 approximately solves these by polynomial time LSMR: optimize c1 + c2 + c3 + ... + cn subject to: c1*x1 + c2*x2 + ... + ck*xk + ... + cn*xn = p d1*x1 + d2*x2 + ... + dk*xk + ... + dn*xn = q c1 >= 0, c2 >= 0, ..., cn >= 0 di are some functions or permutations of ci 1214.4 A random 20523 bits integer has been factorized (small prime factor 7 and a huge factor) in less than 15 minutes which is the largest integer factorized by NeuronRain so far. Advantage of Factorization as proof of work is the parametrizable complexity - Hardness of or Time consumed by Proof of Work can be configured by choosing integers of low prime factors (from number theoretic estimates - Hardy-Ramanujan,Prime Number Theorem,Baker-Harman-Pintz,Cramer,Zhang ray shooting queries) - Integers of low prime factors tend to increase hardness of factorization and viceversa:
Krishna iResearch - NeuronRain - K.Srinivasan
More... over 4 years ago
Section 1213 updated
Krishna iResearch - NeuronRain - K.Srinivasan
More... over 4 years ago
+--------------------------------------------------------------------------------------------------------------------------------------------- +1213. (THEORY and FEATURE) Economic merit - Neuro Cryptocurrency mint - Commodity linked Neuro Fictitious Cryptocurrency Proof of Work Algorithm - related to 164,165,488,489,499 and all sections on Economic merit, Neuro academic-use fictitious cryptocurrency, Economic networks, Towers of Hanoi problem - Ordered Histograms, Set Partition Analytics, Earth mover distance, Psuedorandom generators, NP-complete problems, P Versus NP, EventNet consensus and Hyperledger, Samuleson-Stolper theorem, Invisible Hand - Labour theory of value, Gold Standard - 8 February 2022 +---------------------------------------------------------------------------------------------------------------------------------------------
Krishna iResearch - NeuronRain - K.Srinivasan
More... over 4 years ago
1212. (THEORY and FEATURE) GIS Urban Sprawl Analytics - Population count computation from HRSL Map Legend - 7 February 2022 - related to 1174,1175,1179,1194,1211 and all sections on GIS Urban Sprawl Analytics, Merit of Large Scale Visuals, Gridded Population Estimators (SEDAC GPW4 and GHSL convex hull estimators), DBSCAN and Neural Network Urban Extent Learners, Computational Geometry, Intrinsic Merit Rankings, R Ricker Population Estimation, Community detection, Social Network Analytics, Space filling by Curves, Unequal Circle and Polynomial packing, Computational Geometry and Polygonal packing by Voronoi Tessellation, Four color theorem, Economic Merit, Computational Economics - GDP estimation from electricity consumption, Jordan Curve Theorem, Weirstrass Theorem
Krishna iResearch - NeuronRain - K.Srinivasan
More... over 4 years ago
--------------------------------------------------------------------------------------------------------------------------------------------- 1212. (THEORY and FEATURE) GIS Urban Sprawl Analytics - Population count computation from HRSL Map Legend - 7 February 2022 - related to 1174,1175,1179,1194,1211 and all sections on GIS Urban Sprawl Analytics, Merit of Large Scale Visuals, Gridded Population Estimators (SEDAC GPW4 and GHSL convex hull estimators), DBSCAN and Neural Network Urban Extent Learners, Computational Geometry, Intrinsic Merit Rankings, R Ricker Population Estimation, Community detection, Social Network Analytics, Space filling by Curves, Unequal Circle and Polynomial packing, Computational Geometry and Polygonal packing by Voronoi Tessellation, Four color theorem, Economic Merit, Computational Economics - GDP estimation from electricity consumption, Jordan Curve Theorem, Weirstrass Theorem --------------------------------------------------------------------------------------------------------------------------------------------- 1212.1 GISUrbanSprawlAnalytics.py has been updated to compute population count from a Legend Dictionary through additional argument "legend" to urban_sprawl_from_segments() which is a dictionary of average RGB pixel colors to population range strings of the form "<minimum>-<maximum>" 1212.2 If no legend dictionary is available,population density computation defaults to proportional density based on pixel color having 100000 as maximum. It is upto the invoking function to suitably populate the legend dictionary and mapscale for contour area to square kilometer conversion. An example legend dict for Meta High Resolution Settlement Layer Map testlogs/HRSL_World_NightTimeStreets.jpg has been passed on to urban_sprawl_from_segments() as mapping of RGB averages for color gradings - Yellow1,Yellow2,Yellow3,Orange1,Orange2,Orange3,Red1 - to population ranges. Mapscale could be either set to HRSL legend ratio of 1:147914382 or NeuronRain heuristic of 266.0/854.0 for urban sprawl contour area to square kilometre conversion: legend={235/3:"0-1",245/3:"1-50",255/3:"50-100",362/3:"100-250",372/3:"250-1000",382/3:"1000-2500",510/3:"2500-27057"} mapscale=147914382 #mapscale=266.0/854.0 urban_sprawl_from_segments("testlogs/HRSL_World_NightTimeStreets.jpg",seg9,-1,mapscale,legend) 1212.3 urban_sprawl_from_legends() iterates through legend dictionary to find matching color range for an average RGB value of contour boundary pixel - it is not a key lookup in exact sense, but a range query for color range which contains the average RGB value e.g an average RGB value of 300/3 would be matched by 255/3 - 362/3 bracket. 1212.4 Use of mapscale from HRSL (mapscale=147914382) as multiplicand magnifies the urban sprawl area and population values thus facilitating better ranking. 1212.5 Excerpts of logs from testlogs/GISUrbanSprawlAnalytics.log.7February2022 for area and population estimates of few urban sprawl contours: ('Circumference of Urban Area:', 30.485281229019165) ('Approximate circular area of Urban Sprawl:', 78.5) ('Contour Area of Urban Sprawl:', 60.0) ('Ratio of urban sprawl contour area to image area:', 6.96301247771836e-05) ('Square Kilometer Area of Urban Sprawl:', 8874862920.0) ('Population of Urban Sprawl (from density):', 8874862920.0) ('Population of Urban Sprawl (from area):', 4.8990705329947954e+20) ('Centroid of the Urban Sprawl:', (413, 43)) Contour boundary point: [[387 40]] Contour color: [255 255 255] ('Convex Hull of Urban Area:', array([[[388, 40]],
Srinivasan Kannan
More... over 4 years ago
Sections 1204,1211 updated
Krishna iResearch - NeuronRain - K.Srinivasan
More... almost 5 years ago
Updated Section 1211 and few corrections
Krishna iResearch - NeuronRain - K.Srinivasan
More... almost 5 years ago
+--------------------------------------------------------------------------------------------------------------------------------------------- +1211. (THEORY and FEATURE) GIS Urban Sprawl Analytics - Population Density computation from Contour Pixel color coding - 4 February 2022 - +related to 1174,1175,1179,1194 and all sections on GIS Urban Sprawl Analytics, Merit of Large Scale Visuals, Gridded Population Estimators (SEDAC GPW4 and GHSL convex hull estimators), DBSCAN and Neural Network Urban Extent Learners, Computational Geometry, Intrinsic Merit Rankings, R Ricker Population Estimation, Community detection, Social Network Analytics, Space filling by Curves, Unequal Circle and Polynomial packing, Computational Geometry and Polygonal packing by Voronoi Tessellation, Four color theorem, Economic Merit, Computational Economics - GDP estimation from electricity consumption +---------------------------------------------------------------------------------------------------------------------------------------------
Krishna iResearch - NeuronRain - K.Srinivasan
More... almost 5 years ago
--------------------------------------------------------------------------------------------------------------------------------------------- 1211. (THEORY and FEATURE) GIS Urban Sprawl Analytics - Population Density computation from Contour Pixel color coding - 4 February 2022 - related to 1174,1175,1179,1194 and all sections on GIS Urban Sprawl Analytics, Merit of Large Scale Visuals, Gridded Population Estimators (SEDAC GPW4 and GHSL convex hull estimators), DBSCAN and Neural Network Urban Extent Learners, Computational Geometry, Intrinsic Merit Rankings, R Ricker Population Estimation, Community detection, Social Network Analytics, Space filling by Curves, Unequal Circle and Polynomial packing, Computational Geometry and Polygonal packing by Voronoi Tessellation, Four color theorem, Economic Merit, Computational Economics - GDP estimation from electricity consumption --------------------------------------------------------------------------------------------------------------------------------------------- 1211.1 GISUrbanSprawlAnalytics.py has been updated to compute Population of an urban sprawl contour extent from pixel color coding of the contour and points within the contour. Earlier population density was assumed to be a constant throughout the GIS visual which has been rectified in this code change. 1211.2 There are standard methodologies to find population density from GIS GeoTIFF which project the map pixel coordinates to longitude-latitude by GDAL Geo transformations and lookup the population density of a longitude-latitude from a GIS map service which is multiplied by contour area to get the population density. But the issue with GDAL is it is quite heavy to install and dependency requirements are huge causing C++ compile errors (GDAL is implemented in C++ and there are GLIBC version mismatch errors) and has been abandoned in favour of a simpler alternative. 1211.3 Simpler alternative implemented in this commit: 1211.3.1 Find the Blue-Green-Red color of a pixel in contour boundary of an urban sprawl 1211.3.2 Average of BGR color is divided by 255 to get the ratio of pixel intensity - AverageBGR/255 1211.3.3 Average BGR color pixel intensity ratio is multiplied by highest population density of an urban sprawl in the world to get approximate estimate of population density of an urban sprawl contour (e.g Barcelona,Spain and Los Angeles,The US are the densest in world) - 100000 per square kilometre (Assumption: Pixel colors graded from 0 to 255 map from lowest population density to highest population density range which is the convention followed by GHSL,SEDAC,HRSL) 1211.4 Function urban_sprawl_from_segments() is invoked for Meta-Facebook High Resolution Settlement Layer World Nighttime Streets image - testlogs/HRSL_World_NightTimeStreets.jpg - by setting population density as -1 a flag implying population density is internally computed. 1211.5 Following are the logs, facegraph DOT file and convex hull-contour-segment-facegraph images created - Logs testlogs/GISUrbanSprawlAnalytics.log.4February2022 print the contour boundary point chosen for pixel color computation and the approximate populations of urban sprawl contour areas(not necessarily close to actual populations because area of a pixel may not be 1 square km exactly as contour area to square km conversion is computed by a heuristic training data in 1175, but quite useful for rankings by proportionality): testlogs/GISUrbanSprawlAnalytics.log.4February2022 testlogs/HRSL_World_NightTimeStreets-contourlabelled.jpg testlogs/RemoteSensingGIS/HRSL_World_NightTimeStreets_FaceGraph.jpg testlogs/RemoteSensingGIS/HRSL_World_NightTimeStreets_ImageNet_Keras_Theano_Segmentation_FaceGraph.dot testlogs/RemoteSensingGIS/HRSL_World_NightTimeStreets_contour_segmented.jpg testlogs/RemoteSensingGIS/HRSL_World_NightTimeStreets_convex_hull_segmented.jpg testlogs/RemoteSensingGIS/HRSL_World_NightTimeStreets_segmented.jpg
Srinivasan Kannan
More... almost 5 years ago
Section 1204 updated
Krishna iResearch - NeuronRain - K.Srinivasan
More... almost 5 years ago
Section 1206 updated
Krishna iResearch - NeuronRain - K.Srinivasan
More... almost 5 years ago
Section 1206 updated
Krishna iResearch - NeuronRain - K.Srinivasan
More... almost 5 years ago
1206. (THEORY and FEATURE) People Analytics - Social Network Analysis - Bipartite Social Networks (Matrimonial portals, Employer-Employee Pro fessional networks) - Recommender Systems and Sentiment Analysis of Profiles for Ranking Preferences - related to 583 and all sections on Soci al Network Analysis,Clustering,Bipartite Matching, Gale-Shapley algorithm for Stable Marriage Problem and College Admission Problem, Online Ci tizen Science, Intrinsic Merit ranking of People profiles,Graph Isomorphism and Common Subgraph Problem,Recommender Systems,Unique Identificat ion,Graphical models-Markov Random Fields and Belief Propagation,Hall's Marriage Theorem,Perfect Matching - 31 January 2022
Krishna iResearch - NeuronRain - K.Srinivasan
More... almost 5 years ago
--------------------------------------------------------------------------------------------------------------------------------------------- 1206. (THEORY and FEATURE) People Analytics - Social Network Analysis - Bipartite Social Networks (Matrimonial portals, Employer-Employee Professional networks) - Recommender Systems and Sentiment Analysis of Profiles for Ranking Preferences - related to 583 and all sections on Social Network Analysis,Clustering,Bipartite Matching, Gale-Shapley algorithm for Stable Marriage Problem and College Admission Problem, Online Citizen Science, Intrinsic Merit ranking of People profiles,Graph Isomorphism and Common Subgraph Problem,Recommender Systems,Unique Identification,Graphical models-Markov Random Fields and Belief Propagation,Hall's Marriage Theorem,Perfect Matching - 31 January 2022 --------------------------------------------------------------------------------------------------------------------------------------------- 1206.1 Matching Algorithms for finding complementary pairs of profiles (Match making) in bipartite 2-colored social networks have been studied as Bipartite matching problem and Stable Marriage Problem (Gale-Shapley algorithm). 1206.2 Contrary to homogeneous social networks of Facebook and Twitter, match making social networks consist of 2 sets of vertices (Brides-Grooms,Employee-Employer,Male-Female) and profiles in each set have ranking preferences for their suitor in the other set. 1206.3 k-Nearest Neighbour Clustering of matrimonial profiles has been investigated as a Marriage Recommender Systems engine for suggesting profile vertices of closest match in set V2 for a profile vertex in set V1 of a bipartite social network G=(V1 union V2,E) - Reference 1206.8. 1206.4 Recursive Gloss Overlap Textgraphs of 2 Profile texts from sets V1 and V2 could be matched by Graph similarity measures - Graph Edit Distance,Netrd Graph Distance measures,Graph Isomorphism (Full-Subgraph-Approximate),Common Subgraph Algorithms - for percentage match of the profile textgraphs in bipartite sets V1 and V2 which could be a preliminary recommender systems screening criterion. 1206.5 Unique Identification of a profile in match making is implicit necessity without stating to rule out bigamy. Gale-Shapley Stable Marriage Algorithm discovered in 1962 and awarded Nobel Prize in Economics 2012, has found variety of applications including National Resident Matching Program (https://www.nrmp.org/ for matching healthcare professionals to positions) and states: For equal number of men and women having ranking preferences, it is always possible to find a stable marriage match. A match of two vertices from sets V1 and V2 respectively is said to be stable if there is no pair (A,B) from sets V1 and V2 who prefer each other over their already matched partners. But Gale-Shapley algorithm is heavily tilted in favour of women who are allowed to ditch a current provisional partner found in previous round and choose a better man in next round and thus could be protested as asymmetric, unethical taint for marital sanctity. Thus Gale-Shapley algorithm may not be applicable in its exact sense to ethical matrimonial matching. Runtime of Gale-Shapley algorithm is O(n^2) for n men and n women. Hall's Marriage Theorem is another result which stipulates a condition for unique happy marriage: For two sets of n men and n women, and for subset I of n women, of Ai is the set of men i-th woman in set I prefers to marry, Sum|Ai| for all i in I >= |I| 1206.6 Ranking preferences of a profile required as a preprocessing step by Gale-Shapley algorithm earlier is approximately a Sentiment analysis problem of the profile text or inferring expectations of profile vertices from text. An example matrimonial advertisement profile text of author in SentimentAnalyzer-BipartiteMatching-SocialNetworks.txt (Source advert JPG - Print Media - 2017: testlogs/SentimentAnalyzer-BipartiteMatching-SocialNetworks.IMG_20171111_135651.jpg) has been Deep-learnt to a Recursive Gloss Overlap TextGraph and analyzed for Empath and SentiWordNet sentiment by SentimentAnalyzer.py Belief Propagation and Markov Random Fields Graphical Model sentiment analysis functions which treat the textgraph as Graphical model having sentiment score potentials. Some small changes have been made to potential computation (multiplying by 2 for magnifying the scores, fliter object of sentiwordnet synsets cast to list) 1205.7 Excerpts of logs testlogs/SentimentAnalyzer.BipartiteMatchingSocialNetworks.log.31January2022 print the positivity,negativity and objectivity scores of the profile text - sample profile text is overwhelmingly objective from SentiWordNet and Empath sentiment analysis: ============================================================================== Sentiment Analysis (Applying SentiWordNet to the tokenized text) of the text ============================================================================== Positivity = 0.5 Negativity = 0.125 Objectivity = 12.375 ======= Empath: ======= RGO Markov Random Fields - Empath positivity: 0.0 RGO Markov Random Fields - Empath negativity: 0.0 RGO Markov Random Fields - Empath objectivity: 1478.6871718142365 1206.8 Previous sentiment scores of a profile could be essential ingredients of a larger match recommender systems engine which clusters profiles of similar attitudes by their sentimental score euclidean distances or matches suiting profile vertex pairs (either by Gale-Shapley algorithm or Hall's marriage theorem) which are ranked by sentiment polarity.
Srinivasan Kannan
More... almost 5 years ago
1205. (THEORY and FEATURE) Computational Geometric Factorization - Matplotlib plot of benchmarks - 300 integers from 100 to 400 and 20 integers from 500 to 520 - related to 1159 and all sections on Computational Geometric Factorization, Neural network learning of Factors by SAC=TC=NC=AC equivalence, Planar Point Location - 30 January 2022
Krishna iResearch - NeuronRain - K.Srinivasan
More... almost 5 years ago
--------------------------------------------------------------------------------------------------------------------------------------------- 1205. (THEORY and FEATURE) Computational Geometric Factorization - Matplotlib plot of benchmarks - 300 integers from 100 to 400 and 20 integers from 500 to 520 - related to 1159 and all sections on Computational Geometric Factorization, Neural network learning of Factors by SAC=TC=NC=AC equivalence, Planar Point Location - 30 January 2022 --------------------------------------------------------------------------------------------------------------------------------------------- 1205.1 New source file python-src/DiscreteHyperbolicFactorizationUpperbound_TileSearch_Optimized_MultipleIntegers.py has been committed to iterate multiple integers sequentially over a range and time their computational geometric factorization durations by invoking python-src/DiscreteHyperbolicFactorizationUpperbound_TileSearch_Optimized.py within subprocess.call() 1205.2 NeuronRain Spark MapReduce implementation of Computational Geometric Factorization NC-PRAM-BSP algorithm by local binary search of rasterized hyperbolic arc bow segments is in complexity class Deterministic MRC, a software limitation. Time durations of computational geometric factorization of 300 integers from 100 to 400 and 20 integers from 500 to 520 are plotted against theoretical time of O(logN*logN) for depth-1 non-uniform Nick's Class circuit. 1205.3 python-src/DiscreteHyperbolicFactorizationUpperbound_TileSearch_Optimized_MultipleIntegers.py takes following commandline arguments in the order - minimum integer of the range,increment,depth of non-uniform NC circuit=exponent k of O((logN)^k),constant of Big-O and exponent of logN in the plot. 1205.4 It is hereby clarified that motive of Computational Geometric Factorization algorithm is to devise a polylogarithmic time integer factorization mechanism or a logdepth bounded fanin Nick's class circuit implemented by a cloud computing DMRC simulation of multicore-PRAMs which evidences existence of factorization algorithms in NC and not proving the lowerbound for factorization (e.g possibility of factorization by planar point location being present in a subset of NC cannot be ruled out - sublogarithmic time - since there are sequential sublogarithmic planar point location algorithms) which are two entirely different notions. That is precisely the reason for (ab)use of BigO notation throughout instead of Omega. 1205.5 Equivalence of SAC=AC=NC=TC implies Nick's class parallel algorithms are learnable by Threshold circuits of majority gates or Neural networks - iterated factorization of sequential set of integers might serve as priors for further factorization. Example: Factorization of 39 as 39 = 13*3 might be a prior by substitution for factorization of 117=3*39=3*3*13 1205.6 Two matlibplots python-src/DiscreteHyperbolicFactorizationUpperbound_TileSearch_Optimized_MultipleIntegers1.jpg (for integers 100 to 400,logdepth=1,BigOconstant=1,log exponent in plot=2) and python-src/DiscreteHyperbolicFactorizationUpperbound_TileSearch_Optimized_MultipleIntegers2.jpg (for integers 500 to 520,logdepth=1,BigOconstant=1,log exponent in plot=2) have been committed to AstroInfer repositories. 1205.7 Both the graphics depict the k*logN (k=1) actual runtimes of factorization (blue) far upperbounded by theoretical O((logN)^2) (red). Depth of circuit (parallel RAM time) has been differentiated from exponent of logarithm for graphic illustration (though both have to be equal by definition of NC). Previous runtime plots are for finding all factors (with and without multiplicity - Big Omega and Little Omega) which is a harder problem than finding just one factor.
Krishna iResearch - NeuronRain - K.Srinivasan
More... almost 5 years ago
1204. (THEORY and FEATURE) People Analytics - Social Network Analysis (Professional) - Chaotic Hidden Markov Model of Career Transition - some corrections to Viterbi Trellis computation, JSON persistence and PageRank Model of Career Transition - 27 January 2022 - related to 14,365,711,753,876,1127,1128,1129,1130 and all sections on People Analytics, Intrinsic Merit Versus Fame Equilibrium, BKS Conjecture, Stability and Noise Sensitivity of Interview LTF/PTF,Peres Theorem,Unique Identification,HR Analytics and Automated Recruitment,Question-Answering,Majority Voting,Condorcet Jury Theorem Variants,Margulis-Ruzzo Thresholds,Novelty Detection
Krishna iResearch - NeuronRain - K.Srinivasan
More... almost 5 years ago
-------------------------------------------------------------------------------------------------------------------------------------------- 1204. (THEORY and FEATURE) People Analytics - Social Network Analysis (Professional) - Chaotic Hidden Markov Model of Career Transition - some corrections to Viterbi Trellis computation, JSON persistence and PageRank Model of Career Transition - 27 January 2022 - related to 14,365,711,753,876,1127,1128,1129,1130 and all sections on People Analytics, Intrinsic Merit Versus Fame Equilibrium, BKS Conjecture, Stability and Noise Sensitivity of Interview LTF/PTF,Peres Theorem,Unique Identification,HR Analytics and Automated Recruitment,Question-Answering,Majority Voting,Condorcet Jury Theorem Variants,Margulis-Ruzzo Thresholds,Novelty Detection -------------------------------------------------------------------------------------------------------------------------------------------- 1204.1 SocialNetworkAnalysis_PeopleAnalytics_ChaoticHMM.py has been updated to persist dictionaries for Career State Machine and Designation classifications to SocialNetworkAnalysis_PeopleAnalytics_ChaoticHMM.json JSON backend and remove the hardcoded values. 1204.2 Viterbi Trellis computation has been corrected to find argmax for every state-observation and append to Viterbi trellis for printing to logs testlogs/SocialNetworkAnalysis_PeopleAnalytics_ChaoticHMM.log.27January2022. Career Transition computed by Chaotic HMM is correctly printed sequentially ignoring parallel tenures: ============================ Chaotic HMM Viterbi Trellis computation ============================ ['Schooling', 'DBHPS', 'PSGTech', 'BaaN-SSAGlobal', 'SunMicrosystems-Oracle', 'Krishna-iResearch', 'Verizon', 'webMethods-SoftwareAG', 'CMI-IIT-IMSc', 'GlobalAnalytics', 'Clockwork-PiQube', 'CloudEnablers'] 1204.3 SocialNetworkAnalysis_PeopleAnalytics_ChaoticHMM.json contains dictionaries for an example career transition (academic and work) of author spanning 40 years (1982-2022) which includes parallel tenure of Open Source Software R&D initiative (Krishna iResearch: 2003-2022) spanning 19 years. Lamda of Chaotic HMM has been increased to 5.0. 1204.4 Unique Identification of a profile is of utmost importance for choosing authentic candidates. Usual Face Recognition and Physique Recognition technologies may not be sufficient unless the career transition and financial history of the profile is authorized by a state issued unique ID (Section 711). 1204.5 Modelling Career Transition by PageRank further refines Markov model earlier by replacing websites by organizations. For a career transition probability column-stochastic matrix CT(N) and Initial state of P, following power iteration converges to stationary distribution: CT(N+1) = CT(N)*P Element (i,j) of CT(N) is the probability of state transitioning from organization i to organization j Stochastic Column Vector P is initialized to random probabilities. 1204.6 Previous PageRank model ranks employee preference of corporate organizations quite similar to website rankings and is holistic and not employee specific. Usual "Best place to work" surveys are mostly intrinsic based on worklife quality metrics. 1204.7 SocialNetworkAnalysis_PeopleAnalytics.py already implements a LinkedIn HR analytics for domain specific dictionary for Intrinsic Measurement of Technical Talent (Lines of Code for Information Technology Domain - Open Source Contribution and Corporate Individual Contribution) - such domain specific intrinsic parameters for measuring talent helps in automated recruitment of talent when both Question-Answering (Contests-Examinations-Interviews) and Majority Voting are prone to failures by human errors - thus in Bounded error complexity classes. Stability condition for BKS Conjecture to be true and Interview as LTF has been derived in Section 365. 1204.8 Influence,Average Sensitivity and Noise Sensitivity of LTFs and PTFs are theoretical formalisms of human errors in Question-Answering (when a question or answer is corrupted by some probability). By Peres' Theorem Noise Sensitivity (NS(delta)) of LTF is upperbounded by 2*sqrt(delta) where delta is probability of flipping each bit (or) Pr(f(x) != f(y)) < 2*sqrt(delta) for x correlated to y by delta-probability flip per bit. From 1204.9 Average Sensitivity of PTF is upperbounded by AS(f) ≤ 2^O(d) · log n · n^(1−1/(4d+2)) for degree-d Polynomial Threshold Function in n variables. PTFs are better models of examinations as their coefficients are real-valued and thresholded by sign of polynomial evaluated (Kind of Pass or Fail). Conditions for Human Error in Majority Voting have been analyzed at length by Variants of CJT (p-bias of a voter SAT, Margulis-Ruzzo sharp Threshold at p-bias=0.5) which apply to PageRank as well. Equality of Average Sensitivity/Stability/Sensitivity/Influence of LTF-PTFs and Majority Decision Error of Condorcet Jury Theorem could be a Fame-Merit equilibrium point. Failures of both Question-Answering and Majority Voting by human error would necessitate inenvitably resorting to automated Talent analytics by historic intrinsic merit metrics - Academic qualifications, Total lines of code written, Number of projects and teams managed, Independent Research and Development (Novelty Detection) - which are merit samples of huge durations thus minimizing error. Academic examinations,prone to failure by LTF-PTF sensitivities,as talent yardsticks for students might likewise have to be replaced by merit samples of wider ambit. Class of LTFs is learnable in poly(n/epsilon) or Questions for Answers are learnable in polynomial time (Answer-Questioning or anti-examination where students are supposed to ask questions for given answers)
Krishna iResearch - NeuronRain - K.Srinivasan
More... almost 5 years ago
1203. (THEORY and FEATURE) People Analytics - Merit of Large Scale Visuals - Textual,Topological and Graph Theoretic Face Recognition - Physi que Recognition From Quadrilateral mesh isomorphism and Face Distinguisher from ImageNet predictions - related to 159,660,714,715,720,1185,118 6,1187,1188,1189,1190,1201,1202 and all sections on Topological Face and Handwriting recognition, Product Homotopy, Unique identification, Hom eomorphism and Topological Deformations, Merit of Large Scale Visuals, People Analytics, Computational Geometry, Graph Edit Distance and Earth mover distance, Graph Isomorphism (Exact, Approximate and Subgraph),Graph matching,Graph similarity,Image Search,Quadrilateral Mesh Deformati on and Isomorphism,Maximum Common Subgraph Problem - 24 January 2022
Krishna iResearch - NeuronRain - K.Srinivasan
More... almost 5 years ago