A.I, Data and Software Engineering


Continue training big models on less powerful devices

Table of contentsThe problemThe workaroundCreate a model and check-pointsRelease memory and reload last saved model It would not be a surprise that you may not have a powerful expensive machine to train a complicate model. You may experience the problem of not enough memory during training in some epoch. This article demonstrates a simple workaround for this. The problem Training deep learning...

Create bipartite graph from a rating matrix

bipartite graph from movie lens

Table of contentsThe rating dataTransform the matrix to a bipartite graphGet graph propertiesVisualize the graph As deep learning on graphs is trending recently, this article will quickly demonstrate how to use networkx to turn rating matrices, such as MovieLens dataset, into graph data. The rating data We use rating data from the movie lens. The rating data is loaded into rdata which is a Pandas...

MLP for implicit binary collaborative filtering

multi-layer perception

Table of contentsThe implicit dataThe MLP collaborative filtering modelInputs and EmbeddingConcatenate and MLPData generatorTrain with early stopping callbackWrapping up In this post, we demonstrate Keras implementation of the implicit collaborative filtering. We also introduce some techniques to improve the performance of the current model, including weight initialization, dynamic learning rate...

Create and distribute your python package

python package

Table of contentsPrerequisites:Create folder structureCreating setup.pyGenerating distribution archivesUploading the distribution archives Wrapping up This is a quick guide for create and generate distribution package of your python project so that others can install, import and use in their projects. Prerequisites: You will need the following tools installed in your computer: Python (2.x/3...

Fast uniform negative sampling for rating matrix

Petamind A.I

Table of contentsThe dataThe sampling methodPython Implementation ResultWrapping up Sometimes, we want to reduce the training time by using a subset of a very large dataset while the negative samples outnumbers the positive ones, e.g. word embedding. Another situation when we deal with implicit data. In this case, we may need to populate new data for negative values. This post demonstrates how to...

A.I, Data and Software Engineering

PetaMinds focuses on developing the coolest topics in data science, A.I, and programming, and make them so digestible for everyone to learn and create amazing applications in a short time.

Pin It on Pinterest