Apply your knowledge by building a project

In this project, you'll dive into implementing one of the simplest algorithms to cluster your data: K-Means. Explore it from scratch using only numpy and matplotlib for visualization. We'll stick to the Wine dataset, so at least it will not be boring!

You work as an analyst in a company. The company's HR boss provided you with three datasets. The first two contain information about employees' performance in offices A and B: how much they work, their salaries, the number of their projects, departments, and so on. The third one is an extensive dataset with information on the employees' satisfaction with their jobs, their latest evaluation metrics, and the current status in the company. Your task is to analyze the data and answer some of the HR’s questions.

In this project, we will dive into implementing one of the most popular while simple enough ensemble algorithms — Random forest. Implement the entire algorithm from scratch using numpy. Test it on the titanic dataset from sklearn.

Linear regression is one of the simplest yet powerful tools for finding regularities in data and using them for prediction. It is widely applied both in science and practice. In this project, you will learn how to apply scikit-learn library to fit linear models, use them for prediction, compare the models, and select the best one. You will also learn how to carry out testing for certain issues with data.

Medical researchers have found that if the medicine label contains only text, a person learns only 70% of the information on it. But if a piece of instructions has pictures, a person will learn 95%. In this project, you will explore data about Nobel laureates and apply Python tools for visualization. Start with data cleaning that includes handling missing values and dealing with duplicates. Then proceed to the data visualization. By the end of the project, you will get the graphs showing the main insights for the dataset.

Welcome to a real estate company located in Amsterdam. Your supervisor has given you a task to create a ML model that will predict the price category of a house based on various parameters. The key problem of the task is that there is too much categorical data. In this project, you will learn to work with the input data and to apply ready-to-use machine learning (decision tree) algorithms to your data.

Most galaxies in the Universe are found as highly-dense groups or clusters. The densest and the most populous are galaxy clusters — they may contain 50+ relatively bright galaxies in the area of a few megaparsecs. Smaller aggregations are called groups and compact groups (CGs). Compact groups are of particular interest for studying galaxy merging and the properties of interacting galaxies. This project presents an opportunity to be an astrophysicist. Discover interesting properties of compact groups with the statistical methods of Python. Don't worry if you're completely new to the field of astrophysics — we'll explain all the important concepts and provide you with the tips to help you navigate. But be ready to learn a lot of new information!

Data preprocessing is one of the first steps in the machine learning workflow. The main idea is to transform raw data into a format that machine learning algorithms can easily understand. The predictive performance of a machine learning model highly depends on the input data quality. Thus, it's an absolute must to know how to improve the quality of your input data by removing the features with low predictive value, engineering new ones, and dealing with multicollinearity. With this project, you'll apply these concepts to NBA data to get a high-quality dataset ready to be fed to a linear model!

In this project, we will take a look at the logistic regression algorithm and build a custom binary classifier. The main idea of the project is to implement gradient descent for two different cost functions, devise a method to predict the probability that a given sample belongs to a certain class, and analyze training errors. And last but not least — put your models to the test on a real dataset.

This project uses boosting algorithms to develop and optimize predictive models for medical insurance costs. It encompasses data analysis and feature transformation and provides a basic overview of boosting algorithms. It will guide and support you in implementing and optimizing these algorithms to achieve accurate predictive analytics.

A decision tree is one of the most widely used machine learning algorithms due to its ease of interpretation. This algorithm is similar to the way we make decisions in our daily life.
In this project, you will take a closer look at the algorithm and write it from scratch with the help of Python, NumPy, and Pandas. Teach the model to process categorical and numerical features to make data-based decisions. Implement a decision tree for classification and apply it to a real dataset.

In this project, you are going to explore the main classification algorithms and learn how to find and train the best possible model for the classification of handwritten digits. You will need to process a dataset that includes images of handwritten numbers from 0 to 9. The ultimate goal is to train the model to identify a digit on the picture. Sounds interesting? Let's dive into it!

Let's train a very simple but fully connected neural network! In this project, we'll create the necessary metric functions and use custom feedforward and backpropagation methods and functions, all done by hand. The dataset for this project is Fashion-MNIST – no more boring number recognition.