Updated index.rst and kuja27 mirror

index.rst

@@ -221,6 +221,20 @@ SourceForge - https://sourceforge.net/projects/jaimini/
 
 GitLab - https://gitlab.com/shrinivaasanka/jaimini
 
+Atlassian BitBucket - https://bitbucket.org/ka_shrinivaasan/ (NeuronRain repositories imported as course material supplement t
+o BRIHASPATHI - https://github.com/Brihaspathi - Virtual classrooms)
+
+Bug tracking pages for NeuronRain repositories:
+----------------------------------------------
+SourceForge - NeuronRain Research - https://sourceforge.net/u/ka_shrinivaasan/tickets/
+
+GitHub - NeuronRain Green - https://github.com/shrinivaasanka/Krishna_iResearch_DoxygenDocs/issues
+
+GitLab - NeuronRain Antariksh - https://gitlab.com/shrinivaasanka/Krishna_iResearch_DoxygenDocs/-/issues
+
+(Deprecated) AsFer GitHub issues page - https://github.com/shrinivaasanka/asfer-github-code/issues?q=is%3Aissue+is%3Aclosed).
+
+JIRA Bug Tracking - https://krishnairesearch.atlassian.net/
 
 FAQ
 ---
@@ -369,7 +383,7 @@ isolate the meanings of common pictogram B in three ways by 1) Set intersection
 	47. Finding distance between two tensors of unequal dimensions is a non-trivial problem e.g Computation of distance between two String Syllable Hyphenated 2D Tensors of unequal rows and columns - [["ten"],["sion"]] and [["at"],["ten"],["tion"]] - requires histogram distance measures(Earth Mover Distance,Word Mover Distance,...) because each syllable hyphenated string is a histogram set-partition of the string and each syllable is a bucket.Conventional Edit Distance measure for two strings is 1-dimensional and does not give weightage to acoustics while Earth Mover Distance between two Syllable hyphenated strings is 2-dimensional and more phonetic. In complexity theoretic terms, bound for edit distance is quadratic while Earth mover distance is cubic though there are recent linear complexity EMD and WMD approximation measures - LC-RWMD - Linear Complexity Relaxed Word Mover Distance - https://www.ibm.com/blogs/research/2019/07/earth-movers-distance/ , https://www.ibm.com/blogs/research/2018/11/word-movers-embedding/ . Subquadratic string distance measures if reduced to edit distance imply SETH is false. Closest Pair of N Points algorithm in Computational Geometry is subquadratic O(NlogN) which could be applied to syllable hyphenated String tensor point sets. Towers of Hanoi Problem concerns hardness of moving a single bucket histogram of disks (Animation: http://towersofhanoi.info/Animate.aspx) sorted by descending radii bottom-top to itself preserving sorted order always (Fixed point computation) and only exponential time (2^N - 1 for N disks) algorithms are known for it. Weird counterintuitive fact about Towers of Hanoi: Earth mover distance upon completion of aforementioned exponential number of fixed point moves is 0 - Histogram remains identical after complete move though partial intermediate moves (require minimum 3 single bucket histograms of sorted order) could have Earth mover distance > 0 - or Sequence of EMDs between 3 histograms sinusoidally fluctuates over time for 2^N - 1 moves before eventually reaching 0 (3 histograms unite to 1), a feature strikingly reminiscent of Collatz conjecture. Towers of Hanoi is NP-Hard (Every problem in NP is polytime many-one reducible to Towers of Hanoi) but not known to be in NP (no NP algorithm has been found). Technically, NP-Hard class is unaffected irrespective of P != NP or P = NP - https://en.wikipedia.org/wiki/NP-hardness#/media/File:P_np_np-complete_np-hard.svg. Previous reduction from Towers of Hanoi histograms to EMD sequence is a #P-Complete parsimonious reduction bijection preserving number of solutions - https://en.wikipedia.org/wiki/Parsimonious_reduction#Examples_of_parsimonious_reduction_in_proving_#P-completeness. Towers of Hanoi and other problems in NP-Hard class and #P-Complete problems are thus obvious choice for cryptocurrency proof-of-work (POW) as the hardness (or labour value) of cryptocurrency is insulated from and independent of P!=NP or P=NP. NeuronRain implements Towers of Hanoi (Single Bin Sorted LIFO Histogram) NP-Hard problem as Neuro Cryptocurrency Proof-Of-Work,a harder alternative to NP-Complete ILP Proof-Of-Work.
         48. Graph Edit Distance (GED) is the most fundamental clustering similarity measure which pervades Text-Audio-Visual-People Graph Analytics and Program Analyzers in NeuronRain. Graph Edit Distance generalizes String Edit Distance - every String (and thus Text) is a connected, directed acyclic graph of maximum degree 1 and alphabets are its vertices. Graph Edit Distance between EventNet of a Video and ImageNet ImageGraphs of Images quantifies visual similarity. Graph Edit Distance between weighted automata of two music clips differentiates music (In theory, automata can be checked for equivalence by Table filling algorithm) while GED between Speech-to-Text textgraphs measures audio similarity. Graph Edit Distance between Social Community Graphs, Connections Graph and proper noun filtered (e.g dictionary filter) Textgraphs of People Profiles measures People similarity. Graph Edit Distance between Control Flow Graphs from SATURN, Program Slice Dependency Graphs, FTrace Kernel callgraphs, Valgrind/KCacheGrind/Callgrind userspace callgraphs identify similar codeflow and malwares. While Graph Isomorphism finds similar graphs by vertex relabelling (Exact Graph Matching), Graph Edit Distance generalizes to dissimilar graphs (Inexact Graph Matching).
         49. Transformers are recent advances in Text analytics - NeuronRain Textgraph implementations for Recursive Lambda Function Growth and Named Entity Recognition extend transformers to textgraph vertices degree attention for inferring importance of word vertices of textgraphs.A Question-Answering Bot has been implemented in NeuronRain which takes natural language questions from users and queries wikipedia corpus for answer summary to create a rephrased deep-learnt natural language answer by WordNet walk on edges chosen based on top percentile Transformers Degree attention Query-Key-Values from wikipedia summary textgraph.
-        50. Graphical Event Models (OGEM,PGEM) decipher graph dependency amongst timeseries of real life events (politics,economic and other bigdata streams). EventNet theory and implementation in NeuronRain is a Graphical Event Model for interevent and intraevent actor-model causality. EventNet Tensor Product algorithm for Videos is a Graphical Event Model based on ImageNet for extracting dependencies between frames (Video is a timeseries stream of frames)
+        50. Graphical Event Models (OGEM,PGEM) decipher graph dependency amongst timeseries of real life events (politics,economic and other bigdata streams). EventNet theory and implementation in NeuronRain is a Graphical Event Model for interevent and intraevent actor-model causality. EventNet Tensor Product algorithm for Videos is a Graphical Event Model based on ImageNet for extracting dependencies between frames (Video is a timeseries stream of frames). EventNet Graphical Event Model (GEM) is a 2-dimensional Tensor of interevent and intraevent causalities. Probabilistic EventNet GEM can be learnt from timeseries of events (news articles on socioeconomics and politics) - Learning Bayesian model GEM on example timeseries datasets is described in http://www.contrib.andrew.cmu.edu/org/cfe/simplicity-workshop-2014/workshop%20talks/Meek2014.pdf. Tensor Decomposition of EventNet GEM (decomposition of a Tensor into sum of rank-one product tensors - https://www.kolda.net/publication/TensorReview.pdf) has enormous implications for timeseries causality - real life event causalities in Tensor notation could be classified into linearly independent low rank tensor components. 
 	51. Digital Watermarking overlay of segmented large scale visuals is in a sense a primitive image classifier - vertices of facegraphs of similar segmented images when overlayed on one another are highly superimposed and isomorphic (and thus a measure of similarity) creating a multiplanar graph in which each vertex is a stack - a visual version of ThoughtNet.
         52. Integer Partitions and String complexity measures are related - Every string is encoded in some alphabet (ASCII or Unicode) having a numeric value and thus every string is a histogram set partition whose bins have sizes equal to ASCII or Unicode values of alphabets which partition the sum of ASCII or Unicode values of constituent alphabets of a string. This enables partition distance (a kind of earth mover distance - e.g. Optimal transport and integer partitions - https://arxiv.org/pdf/1704.01666.pdf) between string histograms as a distance measure between strings apart from usual edit distance measures.
         53. Byzantine Fault Tolerance (BFT) has theoretical implications for mitigating faults including cybercrimes in electronic networks and containment of pandemics in social networks modelled by Cellular automaton graphs.

kuja27_website_mirrored/site/kuja27/CV_of_SrinivasanKannan_alias_KaShrinivaasan_alias_ShrinivasKannan.txt

@@ -243,7 +243,7 @@ May 2003 – present – Krishna iResearch (parallel, self-started, not-for-prof
 Working on Research, Design and Development of open source projects – a machine-
 learning, cloud and queue augmented Linux kernel - NeuronRain - started by self, copyleft licensed under GPL v3 .  Premium technical support is available for above opensource codebases. The GitHub repositories implement NeuronRain Green (for cloud and generic datasets),  SourceForge repositories implement NeuronRain Research (for astronomical datasets) and GitLab repositories implement NeuronRain Antariksh (for Drone development). An introductory presentation on NeuronRain based Analytics is at:
 https://github.com/shrinivaasanka/Grafit/blob/master/EnterpriseAnalytics_with_NeuronRain.pdf
-Latest version 2022#07#29:
+Latest version 2022#10#01:
 
 1. NEURONRAIN - Krishna_iResearch_DoxygenDocs – FAQ and Documentation for AsFer, VIRGO, KingCobra, Acadpdrafts and USBmd open source product codebases which are subsystems of Krishna iResearch Intelligent Cloud OS - NeuronRain
 (https://github.com/shrinivaasanka/Krishna_iResearch_DoxygenDocs , http://neuronrain-documentation.readthedocs.io/en/latest/ )
@@ -593,6 +593,29 @@ socket stream
 	2.377 Economic Merit and People Analytics - Mechanism Design of Multiround Majority Voting derived from Simultaneous Ascending
 Auction – for Drone EVM usecase implementation in NeuronRain 
 	2.378 Merit of Large Scale Visuals - Archaeology - Rebus Decipherments of Indus Pashupathi Seals and mining frequent subgraphs from predictions
+	2.379 Astronomy and Cosmology Datasets Analytics - N-Body gravitational accelerations computed for USGS (1900-2012) 8+ magnitude Earthquakes and NOAA HURDAT2 (1851-2012) North Atlantic Hurricanes Datasets
+	2.380 GIS Urban Sprawl Analytics - Facebook (Meta) High Resolution Settlement Layer (HRSL) Population Density Maps - GDAL GeoTIFF to JPEG format Translation
+	2.381 GIS Urban Sprawl Analytics - Facebook (Meta) High Resolution Settlement Layer (HRSL) Population Estimation from GDAL-Rasterio Georeferencing
+	2.382 GIS Urban Sprawl Analytics - Facebook (Meta) High Resolution Settlement Layer (HRSL) and LandSat 9 OLI2-TIRS2 GDAL-Rasterio Georeferencing - Window read
+	2.383 GIS Urban Sprawl Analytics - Facebook (Meta) High Resolution Settlement Layer (HRSL) and Global Human Settlement Layer (GHSL) GDAL-Rasterio Georeferencing - Sampling and Mollweide-EPSG transforms reprojections 
+	2.384 GIS Urban Sprawl Analytics - Facebook (Meta) High Resolution Settlement Layer (HRSL) and Global Human Settlement Layer (GHSL) GDAL-Rasterio Georeferencing - Gini Index of Population and Built-Surface 
+	2.385 GIS Urban Sprawl Analytics - Facebook (Meta) High Resolution Settlement Layer (HRSL) - Population estimation from GDAL-Rasterio Georeferencing
+	2.386 GIS Urban Sprawl Analytics - Polya Urn Urban Growth Model - Revised for N color segments 
+	2.387 NeuronRainApps - Autonomous Driving - Obstacle Lattice from LIDAR Point Cloud Data – Python and C++ lattice walk obstacle avoidance usecases
+	2.388 Computational Geometric Integer Factorization - Python 3.10.4 upgrade - 2232 and 67 bits Quadcore Benchmarks and Mathematica-Pari/GP-FLINT performance numbers comparison
+	2.389 GIS Urban Sprawl Analytics - Polya Urn Urban Growth Model - Weights Learnt
+	2.390 Intrinsic Merit of Music - AI Music Synthesis from Sum of Sinusoids - Signal synthesis from librosa tone() 
+	2.391 Intrinsic Merit of Music - AI Music Synthesis from Polynomials Learnt from training data - PolyFeatures and Carnatic-Hindustani notes support 
+	2.392 Intrinsic Merit of Music - Virtual Piano Implementation and Music Synthesis from 12-notes octave 
+	2.393 Intrinsic Merit of Music - Virtual Piano Implementation - Carnatic Music Synthesis from 12-notes octave 
+	2.394 Intrinsic Merit of Music - Deep Learnt Automata and Music Synthesis - a Boolean Composition and Learning Theory perspective 
+	2.395 GIS Urban Sprawl Analytics - Polya Urn Urban Growth Model - Urban sprawl area computation
+	2.396 GIS Urban Sprawl Analytics - Comparison of Raster Data Bounding Boxes between 2 dates – Chennai Metropolitan Area Sprawl – GHSL GHS SMOD Degree of Urbanization - R2019A and R2022A
+	2.397  Complement Diophantines - Lagrange and Barycentric interpolations 
+	2.398 Intrinsic Merit of Music - Synthesized Bach from training music waveforms
+	2.399 Astronomy and Cosmology Datasets Analytics - Solar system N-Body Pairwise angular separations computed for NOAA HURDAT2 (1851-2012) North Atlantic Hurricanes Datasets
+	2.400 Computational Geometric Integer Factorization - Multiple Integers - Python 3.10.4 upgrade + Spark 3.0.1 Quadcore Benchmarks 
+	2.401 GIS Weather Analytics and Climate Analytics – NASA JPL DE421 Ephemeris N-Body analytics integration and correlation to extreme weather events – Syzygies and angular separations of Solar system bodies
 
 
 AstroInfer is the userspace Big Data Mining and Deep Learning facet of NeuronRain. Initially implemented for mining patterns in astronomical datasets(degrees of astronomical objects viz planets, constellations etc.,) and prediction based on rules and execution of those rules (SourceForge), has been generalized for any dataset (GitHub,GitLab). It is also used in USBmd and VIRGO codebases below for traffic analytics and kernel analytics. Design started in May 2003.