StockWiz

• 

  StockWiz Home Page

• 

  Added Stocks to Portfolio

• 

  Training Model to Predict Stock Price

• 

  Final Stock Portfolio with total profit/loss

Inspiration

The theme came to me when I suddenly got interested in Machine Learning and thought that it would be pretty cool to create my own stock prediction website. So I did exactly that and hope people can enjoy making their own portfolios and playing around with it.

What it does

It is a website that lets you add companies and their stocks along with the quantity you want. Like this, you can make your portfolio of stocks and see their predicted prices the next day along with their present prices. In the end, you can get the final result and find out whether you've made money or not!

How we built it

It uses stock history data time series from ALPHAVANTAGE Api and runs an LSTM (Recurrent Neural Network) using the Tensorflow.js framework. To design and structure the website I used HTML, JavaScript and CSS.

Challenges we ran into

This was the first time I've used Tensorflow.js so it was definitely a challenge to get through by creating the model. Additionally, getting the APIs to work all the time was tricky as some or the other error always popped up.

Accomplishments that we're proud of and what we learned

This was my first hackathon so I'm proud to have built something independently by the end of it! The machine learning model gave a lot of errors initially and eventually, I managed to debug my way out of it. I came to learn many subtleties of JavaScript through this process, and also quite a lot about stocks too. The model showed an accuracy of about 95% on the test set data, which is quite good.

What's next for StockWiz

The next step for StockWiz is to add a multi-user feature where you can compete with your friends in real-time to make the most money with the same budget.

Built With

• alphavantage
• css
• html
• javascript
• tensorflow