StockPrediction-SVM

Support Vector Machine:-

Hyperplane

Seperating plane between two categories of data. The equation of the hyperplane is:-  UW + B
U -> Unknow vector
W -> Perpendicular vector from origin to the hyperplane
B -> Bias variable

Support Vectors

They are the vectors whose removal will affect the hyperplane.

Optimization modules

i) CVXOPT
II) LibSVM

The goal of optimization is to minimize vector 'W' and maximize bias variable 'B'. This optimization problem is known as Convex problem.

Support Vector Regression

It is a regression problem used to find continious value using SVM approach. Support vector regression comes under polynomial regression.

Kernel

It is a funciton that maps lower dimensional data to higher dimensional data.

Conversion of non-linear data to linear data

Non-linear problem is converted to linear one by adding dimensions. That's what kernel does. 

Support vector regression model is used to predict the stock price. There is no much difference between SVM and SVR. Simply, SVM is classification model and SVR is regreesion model.

Microsoft Sentiment Analysis

Sentiment Analysis of Microsoft will be done from Twitter. It will help predict the future stock price based on user's opinions.

Methodology

1) Data Extraction
2) Tweets Preprocessing and Cleaning
3) Extarcting features from cleand tweets
4) Model Building:- Sentiment Analysis
5) Conclusion

Naive Bayes Classifier

Naive Bayes Classifier will be used to determine weather a tweet is negative or positive. It uses an approach of conditional probability. 

Datasets

I have extracted the tweets, likes and retweets of Microsoft related pages. Some of then are:- 1) Microsoft 2)Bill Gates 3)Steve Ballmer 4)Melinda Gates 5)Satya Nadella 6)Windows