This study is about creating a sensitivity classifier model using messages from customers. We have a binary classification problem that categorizes stock sensitivity data as positive or negative. 1 indicates positive sentiment and 0 indicates negative sentiment. The main resource I used in the study is the Python & Machine Learning for Financial Analysis course on Udemy.
The main steps are as follows:
- Importing required libraries(pandas,numpy,seaborn,matplotlib,nltk,gensim,tensorflow)
- Explanatory Data Analysis
- Data cleaning (removing punctuations and stopwords from text)
- Visualization of cleaned dataset and plotting wordcloud
- Prepare the data by tokenizing and padding Building a custom-based deep neural network for sentiment analysis (embedding layer, LSTM network)
- Making prediction and assessing the model performance (confusion matrix)