- SALES PERFORMANCE AND INSIGHTS PROJECT

This project is aimed at enhancing my Power BI skills through a self-driven analysis of a sales dataset sourced from Kaggle. The focus is on generating actionable insights that meet specific business requirements, particularly for management. The objective is to analyze sales figures across various territories and cities, with a special emphasis on current sales performance. The analysis also extends to evaluating the performance of districts and district managers, offering a comprehensive view of sales dynamics across different geographical areas. Through this project, I aim to develop proficiency in data modeling, DAX (Data Analysis Expressions), and creating interactive dashboards in Power BI. The final deliverables will include a series of visual reports that enable management to monitor territory-based sales performance and district management efficiency in real time, driving strategic decision-making.

• Python(Pandas,Numpy,Matplotlib) were used for data EDA, cleaning and preparation

• Power BI, python(Matplotlib, Seaborn, Plotly, or ggplot2) were used for data visualization.

- PROJECT 2: HEART-CARE Deep Learning / Machine Learning Based Cardiovascular Disease Prediction Application

This is the project I did as the final research for my master's degree in July 2023.
This research aims to develop a user-friendly application to predict cardiovascular diseases (CVDs) based on human behavioral features. CVDs, a leading global health concern, are significantly influenced by lifestyle factors.

Methodology:

Dataset: A substantial dataset was used, encompassing a wide range of human behavioral features. Data Preprocessing: Outlier removal, feature selection, and standardization were implemented to ensure data quality and reliability. Machine Learning Models: A diverse set of classifiers were explored, including ensemble models (Random Forest, XGBoost, LightGBM), linear models (Logistic Regression), tree-based models (Decision Tree), distance-based models (K-Nearest Neighbors, SVM), and deep learning models (ANN, TabNet). Hyperparameter Tuning: A wide range of hyperparameters were tested to optimize model performance. Evaluation Metrics: The models were evaluated using train, test scores, accuracy, precision, recall, and F1-scores. Results: TabNet emerged as the top-performing model, achieving nearly 80% train and test scores with promising metrics. Application Development: A user-friendly CVD prediction application was developed using React JavaScript and Tailwind CSS. Conclusion: This research successfully demonstrates the potential of machine learning in predicting CVDs based on human behavior. The developed application offers a valuable tool for individuals to assess their risk and take proactive steps to protect their health.

- PROJECT 3:Google Play Store - Data Analysis Project

This project was undertaken as part of the requirements for the Business Intelligence and Data Analytics module during my Master’s degree program in July 2023.
Google play store is one of the largest and most diversified mobile applications repositories and that hosted more than millions of android base mobile applications under various categories such as entertainment, education, health etc. Millions of mobile users use google play store to download mobile applications perform their day today tasks and google app store has good consumer feedback mechanism, rating and price strategy etc. With the other competitive platforms like Apple play store it’s more challenging for access outside parties and difficult to find balance dataset using Apple play store data. For the tasks of Data Exploration, Data Cleaning and Understanding I get my approach by interpreting using python plotting and numerical libraries such as Pandas, Matplotlib, Seaborn and NumPy which is highly effective in completing tasks efficiently and accurately. I strongly believe that seamlessly visualizing and interpreting data will help to avoid inconsistencies and enhance meaningful insights.
• Tableau used for data visualization.

- PROJECT 4:Generative AI for Mental Health

This project was undertaken as part of the requirements for the Group Module and as a group we focus on use of Generative AI for Mental health.
Exploring the viability, safety, and effectiveness of training generative language models to create empathetic, informative, and contextually relevant supportive text tailored for mental health support. A wide range of mental health issues, including stress, anxiety, and clinical depression, affect millions of people around the world. It is difficult to address these challenges given the stigma associated with seeking mental health services, the limited access to mental health services, and the diversity of human experiences. The ability of generative AI to adapt and generate content provides unique avenues for supporting mental health problems.

Some examples of conversational agents and chatbots include Woebot, Wysa and Tess. Woebot uses cognitive-behavioural and natural language processing techniques to provide users with emotional support and coping strategies and has proven effective in significant reduction in depression (Fitzpatrick et al., 2017). Wysa employs AI to offer emotional support through chat conversations, cognitive-behavioural techniques, and mindfulness exercises and has been utilised in studies to assess the feasibility and effectiveness of text-based conversational agents on users with self-reported symptoms of depression (Inkster et al., 2018). Furthermore, Psychological AI chatbot Tess which provides mental health support, psychoeducation, and reminders through brief conversational exchanges has proven to be engaging and satisfactory and reduces depression and anxiety symptoms in college students self-identified with the disorder Exploring the viability, safety, and effectiveness of training generative language models to create empathetic, informative, and contextually relevant supportive text tailored for mental health support. A wide range of mental health issues, including stress, anxiety, and clinical depression, affect millions of people around the world. It is difficult to address these challenges given the stigma associated with seeking mental health services, the limited access to mental health services, and the diversity of human experiences. The ability of generative AI to adapt and generate content provides unique avenues for supporting mental health problems. Some examples of conversational agents and chatbots include Woebot, Wysa and Tess. Woebot uses cognitive-behavioural and natural language processing techniques to provide users with emotional support and coping strategies and has proven effective in significant reduction in depression (Fitzpatrick et al., 2017). Wysa employs AI to offer emotional support through chat conversations, cognitive-behavioural techniques, and mindfulness exercises and has been utilised in studies to assess the feasibility and effectiveness of text-based conversational agents on users with self-reported symptoms of depression (Inkster et al., 2018). Furthermore, Psychological AI chatbot Tess which provides mental health support, psychoeducation, and reminders through brief conversational exchanges has proven to be engaging and satisfactory and reduces depression and anxiety symptoms in college students self-identified with the disorder Skills: Data Science · Machine Learning · Large Language Models (LLM) · Large Language Model Operations (LLMOps)
• LLM /Machine Learning/ Large Language Model Operations (LLMOps) / LLM Evaluvation methodologies
• LLM Training and Deployment / Azure and Google cloud platforms / Java/React JS/ Tailwind UI/ MYSQL / RESTAPI /

- PROJECT 5:Sales Reports

• Project objective:

  1. Create a customer performance report

  2. Conduct a comprehensive comparison between market performance and sales targets

• Purpose of sales analytics: Empower businesses to monitor and evaluate their sales activities and performance.

• Importance of analyzing sales data: Identify sales patterns and track key performance indicators (KPIs).

• Role of reports: Determine effective customer discounts, facilitate negotiations with consumers, and identify potential business expansion opportunities in promising countries.

• Project objective:

  1. Create Profit and Loss (P&L) reports by Fiscal Year & Months

  2. Create Profit and Loss (P&L) reports by Markets

• Purpose of sales analytics: Evaluation of financial performance, support decision-making, and facilitate communication with stakeholders.

• Importance of analyzing Finance data: Aid in benchmarking against industry peers and previous periods Foundation for budgeting and forecasting.

• Role of reports: Align financial planning with strategic goals Instill confidence in the organization's financial outlook.

• Proficiency in ETL methodology (Extract, Transform, Load).
• Skills to generate a date table using Power Query.
• Ability to derive fiscal months and quarters.
• Establishing data model relationships with Power Pivot.
• Proficiency in incorporating supplementary data into an existing data model.
• Utilizing DAX to create calculated columns.

• Refined understanding of Sales & Finance Reports
• Designing user-centric reports with empathy in mind.
• Optimization of report generation through meticulous fine-tuning.
• Developing a systematic approach to devising a report building plan.