Mpathy

• 

  Landing Page

• 

  REST-API

• 

  PCA-2D Projection Visualization with KMeans Clustering

Inspiration

People use twitter and provide a large amount of data, so when an anomalous tweet is made we can compare new tweets against their previous tweets.

What it does

Pulls training corpus from users twitter feed using twitter API. Sees if new tweets are anomalous compared to previous baseline using Scikit-Learn KMeans Clustering and PCA.

How I built it

Django-REST + Scikit-Learn

Challenges I ran into

Integrating Front/Back End. Data sparsity.

Accomplishments that I'm proud of

UI looks nice. KMeans Clustering kind of clusters in accordance with PCA projection.

What I learned

Django. Scikit-Learn PCA. Seaborn visualization.

What's next for Mpathy

Integrate all components. Deploy it using GCP.

Built With

• django
• python
• react
• scikit-learn
• twitter