Check out neural_nets.ipynb
for the full talk and example code for building a word2vec
Neural Network in Python.
- Ipython notebook, numpy, scipy, pandas, matplotlib, seaborn
gensim
(a C compiler will allow you to train more quickly, though isn't necessary).
You can easily install all of the above with Continuum Analytics' conda - if you haven't heard of it yet, we'd highly recommend taking a look!
The easiest way to install all these packages is the following, once you've gotten conda installed:
conda create --name ds30 --file environment.yaml
We use the following dataset in a few examples. Warning: It's 1.5GB, so sit back and relax while the download happens!
The Google News Model from the "pre-trained" section on this page.
To run this demo, you will need to startup an ipython notebook instance:
ipython notebook
Then go to http://localhost:8888
and click on neural_nets.ipynb
.
You need visit our youtube channel.
This is meant to just give you a brief guided tour of just a few topics in data science.
If you enjoyed this and want to learn more about doing data science in industry, consider applying to be a fellow at The Data Incubator
If you would like to hire data scientists, introduce data science corporate training, or partner to bring The Data Incubator to your country, reach out here.