Skip to content
A machine learning app for practice.
JavaScript HTML Shell
Branch: master
Clone or download
Fetching latest commit…
Cannot retrieve the latest commit at this time.
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
cache
lib
public
screenshots
serverless-fetch
.gitignore
README.md
clean.sh
example.config.json
index.js
package.json

README.md

machine-learning-movies

A Node.js app I'm using to practice machine learning. It will predict movie ratings based on attributes such as year released, genre, etc.

Requires an API key from https://www.themoviedb.org

chart

Forenote

See the example.config.json files for example config formats.

Run

First, place your API key in ./config.json, then:

npm install
DEV=1 node .

In browser, go to http://localhost:8080

There are currently over 350,000 movies in the DB, so the initial non-cached fetching may take several hours since the API has rate limiting of 4 req/sec/IP.

Fetch data faster with AWS Lambda

Since AWS Lambda uses a different IP for each request, we can vastly speed up the fetching with concurrent requests.

At this point, you should have your API key in both ./config.json and ./serverless-fetch/config.json.

cd ./serverless-fetch
npm install
npm install -g serverless
serverless config credentials --provider aws --key MY_AWS_KEY --secret MY_AWS_SECRET
serverless deploy -v

You should get back a URL like so: https://XXXXX.execute-api.us-east-1.amazonaws.com/dev/fetch.

Place that url as the lambda_fetch_url value in ./config.json.

You can’t perform that action at this time.