Objective: The goal here is to create a language detector model based on “The Europarl parallel” corpus using classical techniques of Natural Language Processing and Machine Learning. I will then use this model to detect the language of new sentences taken from Wikipedia. This language detection task falls basically into Text Classification. Read more

Objective: The goal here is to deal with the XOr classification problem that can’t be solved with classical logistic regression models. The XOr, or “exclusive or” problem is a classic problem that can be solved with Artificial Neural Network to predict the outputs of XOr logic gates given two binary inputs. The XOr function returns a true value if the two inputs are not equal and a false value if they are equal. I will apply a such ANN built from scratch to a custom dataset generated for XOr binary classification. Read more

Objective: The goal is to create median house price estimator for Boston city, using a Multiple Linear Regression Model build from scratch, applied on the “Boston Housing Dataset”. Then I will try to improve this estimator, using a Multiple Polynomial Regression Model. I won’t explain theory here, only background basic equations needed for code understanding. Read more

Objective: The goal is to create a salary estimator based on years of experience of the employees of an anonymous company, using a simple Linear Regression Model build from scratch applied on “Salary_Data.csv” Kaggle’s dataset. Then I will compare the performance of the model against the prebuilt Sklearn linear regression model. I won’t explain theory here, only background basic equations needed for code understanding. Read more