Projects
Starbucks Dashboard
(Extensive Data Cleansing, Manipulating, Wrangling using Python, Data Extraction and Modelling using SSMS and Dashboard Development using Tableau)
Key Insights
🎯 Demographics
The majority of Starbucks members are between the ages of 30-45, followed by 18-30.
The gender distribution is relatively balanced, with males slightly outnumbering females.
Income levels are diverse, with a significant portion of members earning between 30k and 60k.
💳 Spending Habits
The 18-30 age group spends the most on average, followed by 30-45.
Females spend slightly more than males across all age groups.
Discount offers are highly effective in driving spending.
📊 Channel Analysis
Email and mobile channels are the most popular for offer delivery.
Web offers have the highest completion rate, followed by email.
Social media offers have the lowest completion rate.
🎁 Offer Analysis
"BOGO" offers are the most popular and have the highest completion rate.
Discount offers are also popular but have a slightly lower completion rate.
Suggestions
Note: This dataset contains information only about transactions from 2013 to 2018.
Based on the dashboard insights, consider the following strategies to increase customer engagement and drive sales:
Target the 18-30 age group with personalized offers and promotions that align with their preferences.
Leverage email and mobile channels to deliver targeted offers and create a more personalized customer experience.
Optimize discount offers to maximize their impact on spending.
Explore new ways to increase social media engagement and drive offer completion.
Analyze customer spending patterns to identify opportunities for upselling and cross-selling.
Implementing these strategies can enhance customer satisfaction, increase loyalty, and drive sales growth for Starbucks.
Airline Review Eda and Sentiment Prediction
(Web Scraping, Extensive Data Cleansing, Manipulating, Wrangling, NLP, CNN, LSBM Classifier)
Scraped the data from a British Website to get the reviews left by verified user for airlines and flights renowned in the world.
Performed Extensive Data Cleaning and Manipulating operations such as TF-IDF vectorization, tokenization, and NLP, to prepare the data for Analysis and Sentiment Prediction using LSBM Classifier, generating predictions with over 97% accuracy.
Restaurant Recommendation System
(Extensive Data Cleansing, Manipulating, Wrangling, NLP, TF-IDF Vectorization, Cosine Similarities)
Performed Extensive Cleansing and Manipulation of the text data of review and ratings of the restaurants provided by the customers, to prepare the data for recommendation and TF-IDF Vectorization using NLP techniques.
After data pre-processing, developed a content-based filtering model, using Co-sin Similarities between the different types of Restaurants, generating 70% accurate results.
Drowsiness Detection System
(CNN, Haar Models, Face Detection, Eye Detection)
Developed a CNN model to detect the state of the eyes being open and shut to process the states of the eyes while detecting Drowsiness.
Used Haar models to identify the face from the video being captured. Later again using the Haar Models identified the eyes from the face to process their state using CNN model.
Developed a python code, to set a count based on the prediction from CNN based on the state of the eyes, to increase the counter when eyes are shut and decrease the counter when the eyes are open.
Later, used an alarm audio file to be played when a certain score of choice is achieved.
Student Database Management System
(Python, MySQL, GUI, RDBMS, OOPS)
Developed Python Code to process student details in the system and track the status of the student application.
Extensively Used OOPs concepts to implement various CRUD operations to develop and access the student database.
Extensively Used Access functions in SQL to provide access to a particular employee list only.
Implemented complex queries along with various features such as providing login, adding students, updating their details, as well as displaying the results for a particular student, and improved the efficiency of the department by 20%.
As a continuation of the project, further procedures can be developed to extract the data on a daily basis and develop reports using python, to extract the employee performance and the flow of student applications.
Shark Tank Data Analysis
(Python, SQL, Data Cleaning, Manipulating, Visualization, R, Power BI, Tableau)
Analyzed and performed EDA on Shark Tank Data set to understand the relationship amongst different entities related to the pitches such as Industry type, Equity Offered, Equity Liquidated, Gender, Number of pitches each year and more.
Answered a few complex queries using SQL, and Power BI and Python Visualizations, shown below -
Select all columns from the 'Shark Tank' table
Count the number of pitches per industry
Count the number of deals per industry where 'Deal' column is 'Yes'
Calculate the conversion rate (percentage of deals) per industry
Count the number of pitches and deals by entrepreneur gender
Calculate the conversion rate (percentage of deals) by entrepreneur gender
Count the number of deals made by each shark
Show the total equity offered and given up per industry, and the times equity liquidated
Also, performed the below mentioned Statistical Tests to understand the effect of different entities on the deal offerings, using the statistical data.
Chi-square test of independence: This test can be used to determine if there is a significant association between two categorical variables, such as Entrepreneur Gender and Deal. The test can help to answer questions such as whether there is a relationship between the gender of the entrepreneur and the likelihood of a deal being made.
T-test or ANOVA: These tests can be used to compare means of continuous variables across different groups. For example, one can test whether there is a significant difference in the equity offered or equity given up by industry type, or whether there is a significant difference in the equity liquidation rate between different groups of sharks.
Regression analysis: Regression analysis can be used to model the relationship between one or more independent variables and a dependent variable. For example, one can investigate the relationship between the equity offered, equity given up, or equity liquidation rate, and the success of the deal.