Data Analysis

Project Overview This project uses a GARCH(1,1) model to forecast the future daily volatility of the Nasdaq Composite Index (^IXIC). The goal is to estimate how much price movement (volatility) to expect in the coming days, which is useful for risk management, trading strategies, and options analysis. The project includes: Downloading daily Nasdaq data from Yahoo Finance Calculating log returns Fitting a GARCH(1,1) model Forecasting 30 days of volatility Comparing model forecasts with historical rolling volatility Wrapping the workflow into a reusable Python class Key Outputs ✅ Nasdaq Price vs. Rolling Volatility (Last 250 Days) Rolling 30-day volatility spikes during market drops. Calm price trends are usually matched by low volatility. ...

🧠 Objective Segment customers of a UK-based online gift retailer using transaction data to uncover behavior-driven groups for targeted marketing. 📦 Dataset Transactions from Dec 2010 to Dec 2011 541,909 records, 8 columns Key columns: InvoiceDate, Quantity, UnitPrice, CustomerID, Country 🔍 Methodology Data Wrangling: Removed canceled orders and missing customer IDs Filtered out negative or zero Quantity and UnitPrice Parsed dates into datetime format Feature Engineering: Built RFM features: Recency: Days since last purchase Frequency: Total purchases Monetary: Total amount spent Preprocessing: ...

This project explores a dataset of Nairobi property listings to uncover patterns in pricing based on location, property type, and features like bedrooms, bathrooms, and house size. The dataset was sourced from Kaggle and cleaned using pandas before performing exploratory data analysis (EDA) and predictive modeling. 🧹 Data Cleaning Summary We performed the following steps: Removed currency symbols and converted prices to integers Renamed columns for clarity (price to price_in_ksh, etc.) Corrected inconsistent entries (e.g., townhuse to townhouse) Handled missing values: Filled missing house_size_sqm using the median by propertytype Dropped duplicate and unused columns Filtered extreme outliers 📊 Exploratory Data Analysis (EDA) 🔹 Average Price by Location ...