0
I Use This!
Very Low Activity

Commits : Listings

Analyzed about 4 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
(*) New intrinsic merit measures based on graph density (regularity lemma) and Bose-Einstein intrinsic fitness have been added as functions and invoked (*) logs have been committed to testlogs/ More... over 9 years ago
Updated AsFer Design Document
Srinivasan Kannan
as K.Srinivasan
More... over 9 years ago
Updated AsFer Design Document
Srinivasan Kannan
as K.Srinivasan
More... over 9 years ago
Updated AsFer Design Document
Srinivasan Kannan
as K.Srinivasan
More... over 9 years ago
(*) Updated AsFer Design Document (*) Updated python-src/software_analytics/DeepLearning_SchedulerAnalytics.py to write in /etc/kernel_analytics.conf (*) logs committed to testlogs/
Srinivasan Kannan
as K.Srinivasan
More... over 9 years ago
Updated AsFer Design Document More... over 9 years ago
First commits for Machine Learning Analytics Driven Linux Kernel Scheduler: More... over 9 years ago
Updated AsFer Design Document
Srinivasan Kannan
as K.Srinivasan
More... over 9 years ago
Updated AsFer Design Document
Srinivasan Kannan
as K.Srinivasan
More... over 9 years ago
Updated AsFer Design Document
Srinivasan Kannan
as K.Srinivasan
More... over 9 years ago
Updated AsFer Design Document
Srinivasan Kannan
as K.Srinivasan
More... over 9 years ago
+----------------------------------------------------------------------------------------------------------------------------------------- +437. (FEATURE-DONE) Approximate SAT Solver update - 24 July 2017 +----------------------------------------------------------------------------------------------------------------------------------------- +(*) Increased number of variables and clauses to (20,20). +(*) Changed SciPy lstsq() to NumPy lstsq() because NumPy has a faster lstsq() LAPACK implementation as against xGELSS in SciPy +(*) Logs for 100 random CNFs have been committed to testlogs/ and Percentage CNFs and Clauses satisfied are as below: +Percentage of CNFs satisfied so far: 58.4158415842 +Average Percentage of Clauses per CNF satisfied: 97.1782178218
Srinivasan Kannan
as K.Srinivasan
More... over 9 years ago
Updated AsFer Design Document
Srinivasan Kannan
as K.Srinivasan
More... over 9 years ago
Updated AsFer Design Document
Srinivasan Kannan
as K.Srinivasan
More... over 9 years ago
---------------------------------------------------------------------------------------------------------------------------------------- Approximate SAT Solver update - 20 July 2017 ---------------------------------------------------------------------------------------------------------------------------------------- (*) Increased number of clauses and variables to 16 and iterated for ~3100 random CNF formulae. (*) Following are the MaxSAT percentages after ~3100 iterations: Percentage of CNFs satisfied so far: 69.1020276794 Average Percentage of Clauses per CNF satisfied: 97.6122465401 (*) Previous values are quite close to 14 clauses and variables iterations done previously.
Srinivasan Kannan
as K.Srinivasan
More... over 9 years ago
Updated AsFer Design Document
Srinivasan Kannan
as K.Srinivasan
More... over 9 years ago
-------------------------------------------------------------------------------------------------------------------------------------------- Recursive Gloss Overlap graph classifier update - 20 July 2017 -------------------------------------------------------------------------------------------------------------------------------------------- (*) Following centrality measures for the definition graph have been printed in sorted order and top vertices are compared: 1) Betweenness Centrality 2) Closeness Centrality 3) Degree Centrality 4) Eigenvector Centrality (PageRank) 5) Core numbers Centrality (k-core dense subgraphs) (*) While first 3 centrality measures are reasonably equal in terms of ranking central vertices, PageRank centrality and k-core centrality are slightly different (*) But all 5 centrality measures reasonably capture the central purport of the text i.e the text on "Chennai Metropolitan Area Expansion to 8848 sqkm" is classified into "Area" which is the topranked central vertex
Srinivasan Kannan
as K.Srinivasan
More... over 9 years ago
+--------------------------------------------------------------------------------------------------------------------- +434. (FEATURE-DONE) Recursive Gloss Overlap Graph Classifier update - Commits 2 - 19 July 2017 +--------------------------------------------------------------------------------------------------------------------- +(*) printed Betweenness Centrality (BC) of the definition graph of a text +(*) BC of a node is the ratio of number of (s-t) shortest paths going through the node to all pairs shortest paths and +thus measures how central a node is to the graph and thus is another basis for classifying a text apart from dense subgraphs.
Srinivasan Kannan
as K.Srinivasan
More... over 9 years ago
--------------------------------------------------------------------------------------------------------------------- (*) Average percentage of clauses satisfied per CNF is printed at the end of 10000 iterations (*) AsFer Design Document updated for Lovasz Local Lemma analysis of least squares CNF SAT solver (*) logs for least square assignment of 10000 random CNFs committed to testlogs/ ---------------------------------------------------------------------------------------------------------------------
Srinivasan Kannan
as K.Srinivasan
More... over 9 years ago
Updated AsFer Design Document
Srinivasan Kannan
as K.Srinivasan
More... over 9 years ago
---------------------------------------------------------------------------------------------------------------------- Approximate CNF SAT Solver Update - percentage of clauses satisfied ---------------------------------------------------------------------------------------------------------------------- (*) This commit prints percentage of clauses satisfied for each random CNF 3SAT formula for both 0 and 1 evaluations (*) logs for ~1000 iterations of random 3SAT have been committed to testlogs/ (*) Interestingly, even failing formula assignments by least squares satisfy more than 75% of clauses of the formula (*) Percentage of formulas satisfied by least squares heuristic is ~70% (*) This is probably a demonstration of complement form of Lovasz Local Lemma (LLL) for MAXSAT which is: if events occur independently with a certain probability, there is a small non-zero probability that none of them will occur (*) Dual of LLL for this MAXSAT approximation is: if each clauses are satisfied with a certain probability, there is a large probability that all of them are satisfied
Srinivasan Kannan
as K.Srinivasan
More... over 9 years ago
Updated AsFer Design Document
Srinivasan Kannan
as K.Srinivasan
More... over 9 years ago
Updated AsFer Design Document
Srinivasan Kannan
as K.Srinivasan
More... over 9 years ago
Updated Requirements.txt
Srinivasan Kannan
as K.Srinivasan
More... over 9 years ago
Updated AsFer Design Document
Srinivasan Kannan
as K.Srinivasan
More... over 9 years ago
Updated AsFer Design Document
Srinivasan Kannan
as K.Srinivasan
More... over 9 years ago
Updated AsFer Design Document
Srinivasan Kannan
as K.Srinivasan
More... over 9 years ago
Updated AsFer Design Document
Srinivasan Kannan
as K.Srinivasan
More... over 9 years ago
Updated AsFer Design Document
Srinivasan Kannan
as K.Srinivasan
More... over 9 years ago
Updated AsFer Design Document
Srinivasan Kannan
as K.Srinivasan
More... over 9 years ago