Stock Prediction By Reading Newspaper Headline Data
We have built an algorithm based on machine learning techniques (PCA, SVM, NN etc) that will read the headlines from the financial magazines and newspapers and will predict the stock prices for that day. This can be helpful for the analysis of the company’ performance. We used PCA as feature extraction method to extract all the keywords from the headlines that would be useful in the algorithm. Furthermore, I used stemming porter and bag of words as the tools for simplifying the word strings and then three classification methods were applied: SVR, NN and Bayesian classifier. Their results were recorded and compared. This was done for ANF, Netflix and IBM. Attached is the graph for ANF. You can contact us to get its source code or to see graphs for the other two companies.
About Me
Having more than 10 years of experience in NLP, machine learning, image processing and computer vision I believe that I can do any task related to ML, deep learning and Data science. I have used almost all algorithms for ML like pca, lbp, lpq, hog, surf, sift etc for features extraction, KNN, decision trees, naïve bayes, random forests, svm, Neural Networks, BERT, Transformers, K-means for classification, clustering and predictive models. I have worked on different Python libraries and frameworks like Tensorflow, pandas, scikit learn,nltk, numpy matplotlib,scrapy, beautifulsoup etc I have used AWS for deploying different algorithms over web and used Flask for making rest Apis. I have also worked on many internet of things related projects in which I integrated machine learning models with iot. I have used node.js and Mqqt for iot related projects. I can make web, desktop and mobile applications related to these fields. In addition, specialized in data labeling and annotation, ensuring high-quality dataset preparation for AI training. I have extensive experience in annotating large volumes of data, including images, text, and videos, tailored to specific project requirements. Using the latest tools and techniques, I provide accurate and efficient data annotation services, supporting a wide range of AI applications. Just inbox your requirements and consider your work done!!
