Generic

Twitter Sentiment Analysis with Gensim Word2Vec and Keras Convolutional Networks

/ Leave a Comment

Fork Word2Vec (https://code.google.com/archive/p/word2vec/) offers a very interesting alternative to classical NLP based on term-frequency matrices. In particular, as each word is embedded into a high-dimensional vector, it’s possible to consider a sentence like a sequence of points that determine an implicit geometry. For this reason, the idea of considering 1D convolutional classifiers (usually very efficient … [Read more…]

Seq2Seq experiment with mathematical expressions

/ Leave a Comment

After reading the article “How to Learn to Add Numbers with seq2seq Recurrent Neural Networks” by Jason Brownlee (that I suggest reading before going on), I’ve decided to try an experiment with more complex expressions like: -(10+5) or 4+ -2, etc. The code (with some extra information) is published on the GIST: https://goo.gl/ZmH6Tf, where there are … [Read more…]

Keras-based Neural Artistic Style Transfer

/ Leave a Comment

Fork I’ve just moved my Keras-based Neural Artistic Style Transfer GIST to a dedicated repository: https://github.com/giuseppebonaccorso/Neural_Artistic_Style_Transfer. Please refer always to it because the GIST is not more maintained. The original post is: https://www.bonaccorso.eu/2016/11/13/neural-artistic-style-transfer-experiments-with-keras/  

OpenAI-Gym Cartpole-v0 LSTM experiment with Keras (Theano)

/ 5 Comments

OpenAI-Gym evaluation page: https://gym.openai.com/evaluations/eval_JxPKNwd1QjaofWkaE4aLfQ. Below there’s the whole Gist page containing the full Python code. It has been developed and tested with Theano/GPU support, but it can easily work with CPU-only support. Any comment or suggestion is welcome! View the code on Gist.

Deep learning, God and Zen emptiness

/ Leave a Comment

It’s quite terrifying discovering how deep learning can work smoothly while performing tasks that even human beings consider difficult. Undoubtedly it’s stunning, extraordinary, but at the same time, it’s almost a terrible discovery. I must confess that I’ve never liked reductionism and I still continue thinking about and “ego” behind the scenes, someone or something, … [Read more…]

Reuters-21578 text classification with Gensim and Keras

/ Leave a Comment

Fork Reuters-21578 is a collection of about 20K news-lines (see reference for more information, downloads and copyright notice), structured using SGML and categorized with 672 labels. They are diveded into five main categories: Topics Places People Organizations Exchanges However, most of them are unused and, looking at the distribution, it’s possible to notice a complete … [Read more…]

Fixed-delay smoothing in HMM with Numpy

/ Leave a Comment

Let’s consider a Hidden Markov Model describing a sequential problem: a system has three internal (hidden) states: [list_items][list_item]Ok (Everything works correctly)[/list_item][list_item]Some issues (not blocking)[/list_item][list_item]Out of order[/list_item][/list_items] However, we can observe only a sensor (globally connected with different sub-systems) which states are represented by three colors (green, yellow and red), representing respectively a normal, partially dangerous … [Read more…]

Theano GPU vs pure Numpy (CPU)

/ Leave a Comment

In this benchmark, I’ve used a Windows 10 Pro 64 Bit computer with Intel Core i7 6700HQ 2.60 GHz with 32 Gb RAM and NVIDIA GeForce GTX 960M. As a programming environment, I’ve used Python 2.7 (Anaconda distribution) and Jupyter. The task is very simple, integrating this expression (simple but effective): [pmath size=24]int{0}{pi}{sin(x)dx} = 2 … [Read more…]

Back to basics!

/ Leave a Comment

This is my brand new blog. I always continue updating my original one (Bonaccorso.org), however I’m going to post here articles related to various IT aspects like: Software Engineering, IT Project Management, OOA/OOD/OOP, J2EE, Enterprise and Mobile Architectures and, of course, funny stuff about my job! Photo credit: fundamentals of industrialization – desktop background wallpaper … [Read more…]