Stock Market Prediction Using Sentiment Analysis and Incremental Clustering Approaches

Sentimental analysis is one of the techniques of Natural Language Processing. It helps us to determine the polarity of the given data: Negative, Positive or Neutral. We know that predicting the prices of Stocks are hard as they are Unstable. Sentimental Analysis helps us to predict the prices of Stocks. The stock prices can be predicted by taking raw data from various social media platforms and these are converted into valuable data by using Sentimental analysis. We observed that we can get more accuracy when we supply more data. DB Scan is an algorithm which is used to form clusters dynamically with the help of eps value. The number of clusters formed will be dependent on the eps value given to the algorithm. And comparing the results of different algorithms tells us which algorithm is best for stock price prediction. Developing the DBSCAN from scratch will gives the flexibility to change the algorithms accordingly and also, we can use the values like centroids in the model. and these results are compared with the algorithms like KMEANS, LSTM, CNN, with the parameters like MAE, MSE, RMSE.