Sunday, December 29, 2024

realSEUDO and evaluation of activation profiles

›
 (This is the part 3 of discussion of realSEUDO). When the activation profiles are generated, there are different ways to scale them that ba...
Sunday, December 22, 2024

LASSO-ISTA-FISTA

›
This is a follow-up to the previous post, talking about the details of the optimization algorithms (in the math sense, i.e. trying to find a...

Triceps and realSEUDO at NeurIPS

›
I've just come back from the NeurIPS 2024 conference. With Triceps listed as my organization. I've done a project with mentoring my ...
Saturday, September 28, 2024

why compilers are called compilers

›
 An interesting tidbit from a recent IEEE Spectrum magazine https://spectrum.ieee.org/from-punch-cards-to-python : The very first "A-0 ...
Wednesday, May 15, 2024

unusual aggregations

›
I've been doing a sizable amount of SQL querying, and it turns out the modern systems have some new and interesting aggregation function...
Tuesday, October 17, 2023

comparing the FloatNeuralNet options

›
I've done a plot of the errors with multiple variations of the FloatNeuralNet options, using the Leaky ReLU as the least common denomina...
Sunday, July 2, 2023

trends

›
Some time ago Linkedin started sending weekly e-mail with the trends for "your kind of jobs". Every week since it's been repor...
Saturday, June 3, 2023

the first actual use of Triceps

›
 I've recently got Triceps used for a real application, for the first time ever, after approximately 13 years. Not really a production u...
Saturday, April 15, 2023

Bayes 29: computation for numeric values (and Python)

›
As promised in the previous installment , here is the how the data science handles the Bayesian inference directly on the numeric values. Re...
Monday, April 10, 2023

Bayes 28: computation by weight revisited

›
I've found out recently how the people in the data science compute the Bayesian inference for values from an arbitrary numeric range (as...
Monday, March 6, 2023

VNC How-to

›
Theoretically speaking, an X terminal can work remotely through a tunnel with ssh -X. But in practice it does that very, very slowly. I don...
Saturday, February 18, 2023

arrested bouncing in FloatNeuralNet

›
It's been a while since I wrote any updates about my experiments with optimization of neural networks. Partially, it's because I...

not MNIST

›
 I've been reading a little more about what other people do with the MNIST recognition. Apparently, doing the transformations on the tra...
Sunday, January 15, 2023

trapeze to btimap for MNIST

›
Continuing with the trapeze-based implementation of the handwritten images that I've described in the previous post, next I've tried...
Sunday, January 8, 2023

trapeze data representation for MNIST

›
 As I've told before , the trouble with MNIST data set is that the test set contains the cases that are substantially different from any...
Thursday, January 5, 2023

unmelting a neural network

›
I've got wondering, could a neural network that experienced a "meltdown" from a too-high training rate be restored? I've g...
Monday, January 2, 2023

optimization 13 - using the gradient sign changes

›
 When I've previously experimented with FISTA momentum descent, one thing that worked quite well was detecting when some dimension of th...

a quick test of theory about MNIST

›
 I've had this theory that the test set in MNIST just contains the digit images that are substantially different from the training set, ...
Sunday, December 25, 2022

optimization 12 - adaptable methods of neural network training

›
 In the last post I've been wondering, why aren't the momentum descent methods widely used for the neural network training. And gues...
Thursday, December 22, 2022

optimization 11 - FISTA momentum descent in FloatNeuralNet

›
Continuing from the previous post, I've had a look at the characters in the test set that have trouble getting decoded. Well, they truly...