Stock-Wizard

Inspiration

I have been really fascinated by Stock market and I have just entered into it but not the Secondary Market but the IPO's. I asked my friend engaged on Stock Market that what sort of automation will make your analysis better and more productive then according to his views and statements I tried to build up and AI model.

What it does

It is an AI model that can be used to Predict High or Low of Shares as well as Closing Stock so that people can effectively predict the results and invest accordingly.

How we built it.

First of all, I grabbed the data from Nepal Stock Exchange and as an Example of model I took an data of NLIC company of Nepal. 2.Then I extracted the data from the website into an CSV file so that it could be loaded into the model as training and prediction testing.

1. I created an VM instance on Google Cloud platform with GPU and run Jupyter Notebook with GPU support so that my training can be fast.
2. Then I trained the model using LSTM neural network and then finally plotted the predicted values and the actual values of the Stock Share using Matplotlib.

Challenges we ran into

The website was made using AJAX so extraction of data using automation was difficult and there was no pagination so I had to manually copy and paste down whole data one by one into excel and later on convert into the CSV file.

Accomplishments that we're proud of

I am really proud that I successfully built an simple model that can actually predict stock prices which will be really handy for people engaged in Stock Market.

What we learned

I learned to make an ML model, Create Instances on Google Cloud Platform, Analyze the Dataset and Plot it respectively.

What's next for Stock-Wizard

I have planned to make an Interactive Interface that allows user to choose High,Low or Closing Price they want to list out using Model and an Interactive options to choose between Different Graphs

Built With

• googlecloudplatform
• jupyter-lab
• keras
• python