Data Scientist | Data Analyst

Technical Skills: Python(Pandas,Sckit-learn,Tensorflow, Keras, Pytorch), SQL, Power Bi, Excel, Tableau ,Data Analyst, Time Series Forcasting, Sentiment Analysis

Education

• B.SC, Information Systems

  Federal University of Technology (August 2023)

• Associate Degree, Computer Craft

  Lagos State Technical College (September 2016)

Work Experience

Data Scientist. Intellidane Cybernetics (September 2020 - November 2022)

• Developed and implemented proprietary algorithmic trading strategies using machine learning models and some statistics indicators.
• I used analytical findings to influence the decision of the strategy to use in implementing proprietary algorithmic trading strategies.
• Developed a machine Learning model that predicts pneumonia severity either high severity or mid severity or low severity using neural network, MLP, Random forest.
• I performed a predictive analysis applying a machine Learning model that predicts several cryptocurrency prices, forcast financial market movement.
• Frequently imported, inspected, cleaned, and transformed data with regard to the business goals of the company using Python. Excel and PowerBi to prepare reports and dashboards.
• Developed a machine Learning model that predicts several cryptocurrency prices using LSTM for the prediction and using Random Forest for the classification of either to buy or sell.

Freelance Data Scientist/Analyst (December 2022 - Present)

• I developed and implemented compelling dashboard using data to tell a story and generate key data points insights For an HR company.
• I develop multiple machine learning models for client( student research) Achievements/Tasks
• Reinforcement Learning-Based Network Intrusion Detection System (https://github.com/akindele22/NetworkIntrusion.git) Developed an AI-powered Network Intrusion Detection System (NIDS) leveraging Reinforcement Learning and the NSL-KDD dataset to detect and prevent malicious network
  activities.
  Key Features:
• Intelligent Threat Detection: Utilized Reinforcement Learning algorithms to identify and respond to network intrusions effectively.
• Dataset Integration: Trained and evaluated the model using the NSL-KDD dataset, ensuring robustness and accuracy.
• Real-Time Monitoring: Enabled dynamic detection of abnormal traffic patterns and potential security breaches.
• Scalable Solution: Designed for seamless integration into existing network infrastructures.
• I developed an image classification using CIFAR-10 datasets.
• Anti-Phishing Chrome Extension with Machine Learning(https://github.com/akindele22/phishing-detector-ext.git)
• Designed and developed an AI-powered Chrome browser extension to detect and flag malicious URLs using a Random Forest Machine Learning model. The extension analyzes URL features in real-time to identify phishing attempts, enhancing online security and user protection.
  Key Features:
• Real-Time Detection: Instantly flags suspicious and malicious URLs.
• Machine Learning Integration: Utilizes a Random Forest classifier for accurate phishing detection.
• User-Friendly Interface: Seamless integration with Chrome for effortless user experience.
• Scalable Backend: Flask API for efficient feature extraction and analysis.
• I interpreted, analyzed and presented the analysis of my data results.
• I developed and designed an interactive dashboard for a software product that shows insights on customer usage.
• Sales Prediction and Analysis Dashboard
  Developed a Sales Prediction and Analysis Dashboard leveraging LSTM for accurate sales forecasting. The project integrates a Flask backend with a Google Colab deployment environment, enabling seamless data preprocessing, model execution, and feature extraction. The dashboard, built with an interactive UI layer, displays sales trends, insights, and future predictions using dynamic visualizations (.
  Key features include:
• Real-time Sales Forecasting: Predict future sales trends with precision.
• Interactive Visualization: Graphical representation of sales data for better decision-making.
• Scalable Deployment: Hosted on Render.

• Collaborative Workflow: Accessible from any device.

• Keystroke-Based Continuous Authentication System Developed an AI-driven continuous authentication system utilizing keystroke dynamics to enhance user security and prevent unauthorized access. Key Features: • Machine Learning Integration: Implemented a Random Forest model for accurate and reliable authentication. • Continuous Monitoring: Enabled real-time analysis of typing patterns for ongoing user verification. • Local Deployment: Optimized the system for seamless performance in a local deployment environment. • Enhanced Security: Improved system resilience against impersonation and unauthorized access attempts.

Projects

Weapon Detection

Developed a machine Learning model built with Tensorflow libraries that detect weapons(knives, matchet ) using inceptionv2, Restnet101,Restnet50.

Crypto trading bot

Developed and implemented proprietary algorithmic trading strategies using simple moving average (200) and exponential moving average(200)

Apple’s Green House Emission

I analyze and design a dashboard that shows how Apple has reduced their emission from 2015 to 2022, I also analyze how this trend compare to their revenue and market capital trend in the same period.

Hybrid Mental Health Support Chatbot

Developed an AI-powered conversational chatbot designed to provide mental health support through empathetic and natural language interactions.
Key Highlights:

• Hybrid Model Architecture: Integrated LSTM for natural language processing and DistilRoBERTa for sentiment analysis.
• Emotion-Aware Responses: Leveraged DistilRoBERTa to analyze user sentiment, ensuring contextually appropriate and supportive replies.
• Conversational Intelligence: Utilized LSTM for advanced NLP modeling, enabling fluid and context-aware conversations.
• Real-Time Interaction: Delivered responsive and engaging user experiences tailored to mental health assistance.

Inventory Management Dashboard

I designed a compelling dashboard on inventory management based on ABC analysis and XYZ analysis in inventory management system. ABC analysis is a method in which inventory is divided into three categories, i.e A,B and C in descending value. The items in the A category have highest value, B category items are of lower value than A, and C category items have the lowest value. XYZ analysis is a way to classify inventory items according to the variability of their demand or derived/forcasted consumption. XYZ analysis can be used to plan material requirements and inventory levels so that waste, production delays, or excessive inventory levels can be avoided. I use DAX to create a measure for the analysis.