Shiva Bajelan's Projects
Belly Button Biodiversity Dashboard is an open-source interactive dashboard that visualises the Belly Button Biodiversity dataset. Built with JavaScript, D3.js, Plotly.js, HTML, and CSS, the dashboard features include a dropdown menu, horizontal bar chart, bubble chart, demographic information display.
The aim of this project is to use the logistic regression mode as a binary classifier to analyse credit card risk. The recommended model helps to predict the high-risk cases. The accuracy, precision, and recall metrics are used to evaluate this model performance.
The goal of this project is to build an ETL pipeline using Python, Pandas, Python dictionary methods to extract and transform the data. Four CSV files will be created and they will be used to create an ERD and a table schema. Finally, the CSV file data will be uploaded into a Postgres database.
This project applies K-means algorithm to group cryptocurrencies based on 24-hour and 7-day price changes. It also investigates the impact of dimensionality reduction using PCA on clustering outcomes.
Repository for course work for UWA Viz bootcamp
This project uses deep learning to solve a classification problem. The dataset was preprocessed and a neural network model was optimized to achieve the target performance. Various techniques were tried to improve the model, demonstrating the power of deep learning models for classification problems.
Performed comprehensive Exploratory Data Analysis (EDA) on Global Significant Earthquake data to uncover patterns, trends, and insights.
The analysis in this project aims to provide insight into the climate patterns of Honolulu in Hawaii and inform decisions regarding the best time to visit and what activities to plan.
This project analyses home sales data using PySpark SQL. It involves creating a temporary table, running queries, and performing caching and partitioning. The final step involves uncaching and verifying the temporary table.
The SQL project involved designing tables to hold data from six CSV files, creating a table schema for each file, importing the data into SQL tables, and performing data analysis. The analysis involved answering various questions about the data, such as listing employee information and department managers and ...
Module 4 challenge
In this project, we looked at Yelp data about restaurants and bars in Perth and performed exploratory data analysis to determine relationships between some different variables. An interactive map is also created giving user the chance to choose their desired criteria from a list.
The purpose of this analysis is to create a binary classification model using different machine learning techniques to predict if an individual with HIV symptoms will be infected with AIDs after receiving a particular treatment after 20 days. The performances for all the five models in this project are compared at the end.
Module 3 challenge
The study involved treating 249 mice with SCC tumors using a range of drug regimens, including Pymaceuticals' drug of interest, Capomulin. Over 45 days, tumor development was observed and measured to compare the performance of Capomulin against other treatments. My task was to generate tables and figures for the technical report of the study.
The goal is to help the editors of a food magazine, Eat Safe, Love, to evaluate the data and assist their journalists and food critics in deciding where to focus future articles. The project aims to provide insights into the ratings data to identify establishments that meet the magazine's criteria for featuring in their articles.
USGS Earthquake Visualisation is an open-source project that provides an interactive map to visualise earthquake data collected by the USGS, highlighting the relationship between tectonic plates and seismic activity. Built with JavaScript, Leaflet.js, D3.js, HTML, and CSS, the project is available on GitHub under the MIT License.
Module 2 challenge, VBA scripting
Module 6 challengeThis project involved using Python and an API to investigate weather trends near the equator by collecting and analyzing weather data. The analysis helped to draw conclusions and provide insights into the factors affecting weather trends in this region.
I used BeautifulSoup and automated browsing to extract information about Mars from two different sources. In Part 1, I scraped titles and preview text from Mars news articles, while in Part 2, I scraped and analysed Mars weather data to gain insights into the planet's climate patterns.