This repository contains examples of how to use Brainome Daimensions to solve various real-life problems. Each directory contains the files needed to reproduce a solution given the Brainome Table Compiler.
Beginner Examples:
Kaggle Titanic -- Kaggle's famous example of predicting survival based on various properties in the passenger list. This notebook is the perfect place to start! For more comments and explanations about the bash outputs, check out the pdf version.
Breast Cancer -- UCI's breast cancer dataset. This notebook is a great example of how to use Daimensions' attribute rank option and offers insight into why attribute ranking is so powerful for understanding data.
Bertrand Synthetic Data -- The dataset was created by Brainome cofounder Bertrand Irissou with a specific rule in mind. This notebook is a test of how well Daimensions works on artificially created data to see if it can find the rule for all data points.
USPS -- The dataset is a multiclass dataset from OpenML of handwritten digits taken from the US Postal Service. Here, we can see the high accuracy of Daimensions on a multiclass dataset.
MNIST -- The dataset is a well-known set of handwritten digits from OpenML. This notebook shows how Daimensions performs on it.
Plotting -- This folder has examples of how to make png visualizations of various dataset predictions in the hidden space.
Spotify -- Extract out features from spotify data automaticaly to indicate a person's personal music preferences with high accuracy.
Advanced Example: Speech/Non-Speech/Music Detector -- An example of a speech/non-speech/music detector for random audio files.
Activity Levels -- How to use tryall.py to try -f NN, -f DT, -f RF, -f RF -rank, -f DT -rank, -f NN -rank
Credit Card Fraud -- Learn how to use -O and -balance for highly imbalanced data
How To Use Soft Probs -- Learn How to use the softprobabilities of our classify function in our compiled python predictors
More to come...
React
Typescript
Django
Laravel
Facebook
Microsoft
Google
Alibaba
D3
Tencent