Tom Kroot Data Portfolio
Welcome to my data portfolio! Here, I showcase my various projects and demonstrate my skills in data analytics, machine learning, and data visualization. Feel free to browse through my portfolio and get in touch with me for any queries or collaboration opportunities.
Stock Price and News Sentiment Correlation
In this project, I used machine learning algorithms to predict the stock prices of a tech company. I was responsible for collecting and cleaning the data, implementing various models, and analyzing the results. The project helped me gain expertise in time-series analysis and predictive modeling.​
​
Covid Data Analytics
In this data analytics project, I analyzed global COVID-19 statistics from Our World in Data, focusing on cases, deaths, and vaccinations. I integrated this data with an SQLite server, where I created various tables and views for detailed analysis. My project involved calculating mortality rates and infection percentages in relation to the total population, as well as identifying the highest infection and death counts by country and continent. I also delved into vaccination data, tracking the rolling count of vaccinations and determining the percentage of the population vaccinated. To effectively present this complex data, I utilized Tableau for visualization, transforming intricate datasets into comprehensible and informative visual formats. Through this project, I was able to apply my data analytics and visualization skills to derive meaningful insights from the COVID-19 data.
​​
Tableau Dashboard​​
Analyzing the Distinction Between Human-Written and Machine-Generated Texts
In this project I focused on differentiating human-written essays from machine-generated ones using a specialized dataset. Employing Python and NLP techniques, including TF-IDF vectorization, I prepared and processed the text data for machine learning analysis. I used a RandomForestClassifier to train a model on a balanced split of training and testing data. The model's performance was outstanding, achieving an accuracy rate of approximately 98.38%, as evidenced by high precision and recall values in both text categories. This project not only highlighted the efficacy of machine learning in text classification but also underscored its potential in applications like content verification and academic integrity.
​
