- created a website which can predict following
- HIGHEST values of a particular company
- LOWEST values of a particular company
- CLOSING values of a particular company
- Algorithms used:
- Decision Tree Regression
- Random Forest Regression
- Support Vector Regression *Data taken from ** https://in.finance.yahoo.com/
- Use pandas-datareader package to connect to yahoo server to fetch dataframe
Python: 3.7 Packages : pandas, pnadas-datareader, sklearn, numpy, requests, django, matplotlib, seaborn
Initiallly, I connected to yahoo server using pandas-datareader package to fetch data of client defined duration(preffered : data of month or more).
Since data fetched is mostly numerical no need to clean data.
I started implementation with Linear Regression which in this case performs below expectations as can be seen in case studies.
After parameter tuning I figured out RandomForest, SVR and Decision Tree performs pretty well
In this step I thought of creating an interactive application for end user. After a lot thought I came up with developing website using
django framework.
Screenshots of website