Tools Used: Python, Pandas, Matplotlib, Seaborn, Scikit-Learn, NLTK/WordCloud
This repository contains my Data Analytics projects completed during my internship at Codveda Technologies.
- Loaded and labeled the Boston House Pricing dataset
- Checked for missing values and duplicates
- Removed price outliers using the IQR method
- Saved a clean version of the dataset for analysis
- Bar chart: Average house price by number of rooms
- Scatter plot: Crime rate vs house price
- Line chart: House age vs average price
- Houses with more rooms tend to cost more
- Higher crime areas have lower house prices
- Newer houses are generally more expensive
Regression Modeling & Clustering ###Task 1: Linear Regression Modeling
- Built a regression model to predict numerical targets based on dataset features.
- Extracted and visualized the model coefficients to determine which features have the strongest impact on predictions.
- Evaluated model accuracy by plotting Actual vs. Predicted values.
- Analyzed Residuals to ensure the model's errors were randomly distributed.
###Task 3: K-Means Clustering
- Applied the K-Means unsupervised machine learning algorithm to segment the data into distinct groups.
- Generated an Elbow Plot to mathematically determine the optimal number of clusters (K).
- Visualized the final K-Means Clusters to highlight distinct groupings within the dataset.
###LEVEL 3: Classification & Natural Language Processing (NLP) ###Task 1: Classification Modeling
- Built classification models to categorize data into discrete classes.
- Generated a Confusion Matrix to assess the accuracy, precision, and recall of the model's predictions.
- Visualized a Model Comparison to highlight the performance differences between various machine learning algorithms.
- Plotted the Predicted Distribution to show the frequency and balance of the predicted classes.
###Task 3: Sentiment Analysis
- Applied Natural Language Processing (NLP) techniques to extract and classify sentiments from text data.
- Visualized the overall Sentiment Distribution to show the proportion of positive, negative, and neutral sentiments.
- Created Word Clouds to highlight the most frequent terms overall, as well as specific keywords driving Positive and Negative sentiments.
Chukwuma Akachi Blessing