Data Analytics

The objective of this project is to conduct a comprehensive analysis of Scottish schools in order to derive valuable insights. The analysis includes clustering and descriptive analysis, taking into account factors such as deprivation rate and total number of pupils. Additionally, an interactive map of Scotland has been developed to visually represent the location of each school. The clustering model has categorized local authorities into three distinct clusters based on average pupil count and deprivation score. The map highlights areas predominantly occupied by schools facing high levels of deprivation. The aim is to assist charities or non-governmental organizations (NGOs) involved in projects supporting these school pupils. All data used in this analysis is publicly available and sourced from the Scottish Government website. The Postcodes.io API was utilized to gather latitude and longitude coordinates for each school based on their respective postcodes. The codes for clustering, visualizations, map generation, and API application can be found in my GitHub repository. You are welcome to access and utilize the repository’s contents for personal or commercial purposes without seeking my consent. I hope you find this information enlightening and enjoyable to read. ...

Project description Language: Python Working file: Microsoft Visual Studio Code Project type: Web Analytics Companies receive traffic to their website from various resources. The more a company learns about the visitors’ traffic to its website, the better it understands the reasons behind their travel to the website. Besides the number of visitors to a website, it is important to learn from which sources those people arrive at the website. This can help the firms to see which external social account or network campaign is prone to bringing new customers. Furthermore, a company can also see the patterns in the behaviour of the visitors on different pages. By using various page analytics techniques, successful as well as problematic pages can be detected. ...

This project is carried out by me independently. The dataset is obtained publicly from Kaggle. All codes and their explanations are stored in my GitHub repository. Project description Language: Python Libraries: sklearn, pandas, numpy, matplotlib, seaborn IDE: Jupyter notebook Project type: Machine learning, Unsupervised learning, K-Means clustering FIFA 23 is a football video game created by Electronic Arts (EA). It became the best-selling football video game in Christmas UK retail charts. According to EA statistics, the game contains more than 700 teams with over 19,000 football players, playing in at least 30 football leagues. The data used in this project is taken from Kaggle. The objective of this project is to classify the players into various segments. ...

This project is carried out by me under a final assignment for one of my courses at the University of Edinburgh. The project received an A grade. Project description Language: Python Libraries: requests, pandas, numpy, matplotlib, seaborn, os, math, time, datetime, json IDE: Microsoft Visual Studio Code, Jupyter Notebook Project type: Data analytics, Web scraping, API Companies House is an agency formed by the British Government to maintain the registration of all the companies in the UK. It maintains a database that stores the information of the registered companies. Each company has features such as company name, Standard Industrial Classification (SIC) code, creation date, cessation date (if ceased operating), company board and shareholder info, etc. By using an API, it is possible to scrape the data from that database by using various specifications. ...

Project description Language: Python Working file: Jupyter notebook In 2018, former President of South Africa, Jacob Zuma, established a commission of enquiry in state capture, known as The Judicial Commission of Inquiry into Allegations of State Capture, Corruption and Fraud in the Public Sector including Organs of State, or simply the Zondo Commission. The commission collected one exabyte of evidence, and on 22 June 2022 released its final report. The reports are available publicly in this link. ...