Chaitanya Kaul

Aim of the problem was to predict the price of the car. Performed Feature engineering, exploratory data analysis, handling of categorical data, Hyperparameter tuning. Compared the results given by various the Regression models. Random Forest Regressor gave the best results. Used the Flask framework for web application and Deployed the model on Heroku.

Regression problem statement, Collected the data by performing web scrapping. Performed feature engineering, feature selection, hyperparameter tuning, exploratory data analysis, Compared the results given by various Machine Learning models and ANN. Best RMSE score given by Random Forest Regressor which is equal to 38.85. Used the Flask framework for web application and Deployed the model on Heroku and Google Cloud Platform.

Its a Regression Problem where the aim of the project is to predict the price of the flight ticket.
Performed feature engineering, feature selection, Hyper parameters tuning using randomized search CV.

Its a deep learning problem where aim is to classify whether a person has Malaria or not.
We use transfer learning technique(VGG 19) for the classification of our image dataset.
Accuracy achieved 97.6 %.

The aim of the problem is to classify whether the news is fake or not.
Its a binary classification problem.
First we are preprocessing the text data using NLP concepts and then
We are implementing this problem using Recurrent Neural networks i.e LSTM & Bi directional LSTM.

Creating and comparing accuracies of different machine learning classification models to classify whether a transaction is fraud or not.
Performed feature engineering, handeling imbalanced dataset.

We have a data which classified if patients have heart disease or not according to features in it. We will try to use this data to create a model which tries predict if a patient has this disease or not. We will use logistic regression (classification) algorithm.

Built a Movie recommendation system by performing Demographic filtering, Collaborative filtering and Content Based filtering.
used Cosine similarity to find similarities between two movies.

Classifying messages as spam or not spam by applying Natural language processing and Using Binary classifiers.

Its a binary classification problem where aim is to classify whether an individual will get a loan from bank or not.
We analyse several parameters in this problem statement , perform data analysis, feature engineering , detection of outliers , looking at feature correlation and finally modelling.
In modelling we compare the accuracies of 4 different supervised machine learning models.

Its a binary classification problem where aim is to classify whether the price of the stocks will increase or decrease based on the headlines in the newspaper.
First we perform text preprocessing.
We use Bag of words and Tf-IDF techniques for converting text into vectors.
Finally we use Ramdom Forest and Naive Bayes classifier for the classification purpose and and compare the different models.