Prediction Using Supervised ML

Project Overview

This project is part of the Graduate Rotational Internship Program at The Spark Foundation. It consists of two main tasks:

1. Prediction Using Supervised Machine Learning
2. Species Segmentation Using K-Means Clustering

Author

Oluwashina Dedenuola

Libraries Used

import pandas as pd
import matplotlib.pyplot as plt
import numpy as np
import seaborn as sns
from sklearn.linear_model import LinearRegression
%matplotlib inline

Dataset

The dataset used in this task was provided by the Spark Foundation.

Steps

1. Data Loading: The data is imported and displayed successfully.
2. Data Analysis: Descriptive statistics of the data are generated.
3. Data Visualization: A scatter plot is created to show the relationship between hours studied and scores obtained.
4. Regression Model: A Linear Regression model is created to predict scores based on hours studied.
5. Predictions: The model is used to make predictions for given hours of study.
6. Visualization of Regression Line: The regression line is plotted on the scatter plot to visualize the predictions.

Results

The model shows a high R-squared value of approximately 0.953, indicating a strong relationship between study hours and scores.

Task 2: Species Segmentation Using K-Means Clustering

Libraries Used

import pandas as pd
import numpy as np
import matplotlib.pyplot as plt
import seaborn as sns
from sklearn.cluster import KMeans

Dataset

The dataset used for this task is the Iris dataset, which contains measurements of various features of different Iris species.

Steps

1. Data Loading: The data is loaded from a CSV file provided by the Spark Foundation.
2. Data Mapping: The species names are mapped to numerical values for clustering.
3. Data Visualization: Scatter plots are created to visualize the distribution of data points based on sepal length and width.
4. Clustering: K-Means clustering is performed, and the optimal number of clusters is determined using the Elbow method.

Results

The clusters are visualized, providing insights into the distribution and segmentation of different Iris species.

Conclusion

This project demonstrates the application of supervised and unsupervised machine learning techniques in predictive analysis and clustering. It provides a practical understanding of regression analysis and clustering methods.

Future Work

Further improvements could include:

• Testing additional algorithms for better accuracy.
• Applying feature scaling techniques for better clustering results.

Acknowledgments

• The Spark Foundation for the opportunity to participate in the Graduate Rotational Internship Program.