SoftNexis Internship Projects

Overview

This repository contains internship projects focused on data preprocessing, exploratory data analysis, statistical testing, and time series forecasting using Python. These projects demonstrate practical data science skills including data cleaning, visualization, hypothesis testing, predictive modeling, and business insight generation.

---

Project 1: Data Cleaning and Preprocessing

Description

This project focuses on improving data quality by cleaning and transforming raw datasets for further analysis.

Tasks Performed

• Dataset inspection
• Handling missing values
• Removing duplicates
• Correcting data types
• Standardizing formats
• Removing irrelevant columns
• Exporting cleaned data

Technologies Used

• Python
• Pandas
• Jupyter Notebook

Outcome

Produced a clean and structured dataset ready for analysis and modeling.

---

Project 2: Exploratory Data Analysis (EDA) – Titanic Dataset

Description

Performed exploratory analysis on the Titanic dataset to identify factors influencing passenger survival.

Tasks Performed

• Data inspection
• Missing value analysis
• Statistical summaries
• Correlation analysis
• Outlier detection
• Data visualization
• Hypothesis testing

Technologies Used

• Python
• Pandas
• NumPy
• Matplotlib
• Seaborn
• SciPy

Key Findings

• Female passengers had higher survival rates.
• First-class passengers showed better survival chances.
• Passenger class significantly influenced survival.
• Fare and passenger class exhibited a strong relationship.

Output

• Visualizations
• Statistical summaries
• Analytical insights and recommendations

---

Project 3: Time Series Analysis and Forecasting

Description

Analyzed historical airline passenger data to identify trends, seasonality, and forecast future passenger demand.

Tasks Performed

• Data preprocessing
• Trend visualization
• Seasonal decomposition
• Moving average smoothing
• SARIMA forecasting
• Model evaluation using RMSE

Technologies Used

• Python
• Pandas
• NumPy
• Matplotlib
• Statsmodels
• Scikit-learn

Key Findings

• Strong upward growth trend observed.
• Significant yearly seasonality detected.
• SARIMA effectively modeled trend and seasonal components.
• Forecasts can support capacity planning and resource allocation.

Model Performance

RMSE: 21.17

Generated Visualizations

• Time Series Plot
• Seasonal Decomposition
• Moving Average Trend
• Forecast vs Actual Comparison

---

Project 4: A/B Testing for Website Conversion Optimization

Description

This project evaluates whether a redesigned website improves user conversion rates compared to the existing website using statistical hypothesis testing.

Tasks Performed

• Dataset creation and preprocessing
• Conversion rate analysis
• Two-proportion Z-test
• Confidence interval estimation
• Data visualization
• Business impact assessment

Technologies Used

• Python
• Pandas
• NumPy
• Matplotlib
• Statsmodels

Dataset Summary

Group	Visitors	Conversions	Conversion Rate
Existing Design	10,000	750	7.5%
New Design	10,000	900	9.0%

Results

Z Statistic : 3.8552
P Value     : 0.000058

Key Findings

• The new design achieved a 9.0% conversion rate.
• The existing design achieved a 7.5% conversion rate.
• Conversion rate improved by 20%.
• Statistical testing confirmed the improvement is significant.

Business Recommendation

Deploy the new website design to improve conversion performance and customer engagement.

---

Technologies Used

• Python
• Pandas
• NumPy
• Matplotlib
• Seaborn
• SciPy
• Statsmodels
• Scikit-learn
• Jupyter Notebook

---

Conclusion

This repository demonstrates practical applications of:

• Data Cleaning and Preprocessing
• Exploratory Data Analysis (EDA)
• Statistical Hypothesis Testing
• A/B Testing
• Time Series Forecasting
• Data Visualization
• Business Insight Generation

These projects showcase end-to-end data analysis workflows and the ability to transform raw data into actionable insights using Python and modern data science tools.