SHALAKA DINESH SONJE

IT Professional with 1 Year and 8 Month Plus Experience in Python and Data Science.

Pune

An independent and self-motivated professional with excellent Python,Machine Learning and Artificial Intelligence Algorithms able to grow positive relationships with clients and colleagues at all organizational levels.

▪ Imbibe scientific in-depth knowledge and interpersonal abilities.

▪    Quick learning abilities to adapt to current technology and market needs
▪    Experience in machine learning problem solving during projects in current job.
▪    Experience in data science libraries like NumPy and pandas with thorough knowledge of python visualization in Seaborn, plotly and matplotlib.
▪    Proficiency in coding Python for Artificial intelligence including Data science and Machine learning.

JOURNEY

2018

Tech Mahindra
Technical Support Associate, Hinjewadi Phase 3 Pune

Text mining / Sentiment Analysis	Twitter data analysis to check whether the sentiment of customer is positive or not.
Description	1. Connecting to twitter using API. 2. Cleaning the text data to remove all punctuations and special characters. 3. Prepare word cloud using Word cloud function. 4. Defining a function to take a sentence as input and provide a sentiment for it. 5. Splitting the sentence into list of words 6. Iterate every word and check if it is present in the positive or negative word.

Clustering	Apply clustering on customer data and find out groups of prime, Good and average customers
Description	1. Import csv file. 2. Finding the best number of clusters based on the distortion value. 3. Defining the K-Means object for best number of clusters. 4. Running the clustering algorithm on the data set. 5. Plotting scatter values.

Classification	To find out whether a customer is Prime, good or average customer.
Description	1. Import csv file. 2. Exploring data distribution of categorical variables using bar plot. 3. Remove Useless Columns. 4. Missing values treatment. 5. Visualizing the important characteristics of a dataset using bar plot. 6. Sampling the data into two parts: Training & Testing. 7. Applying Random Forest Algorithm & find Model Accuracy using f1-score.

Multi Linear regression	To find out how many dealers are required in particular are depending on number of customer and sell in the area
Description	1. Import csv file. 2. Selected required columns as per plot object. 3. Remove Useless Columns. 4. Missing values treatment. 5. Sampling the data into two parts: Training & Testing 6. Multi linear regression algorithm. 7. Measured model accuracy with MAPE & APE.

2018

Tech Mahindra Limited
Technical Support Associate, Hinjewadi Phase 3 Pune

Churn could happen due to many different reasons and churn analysis helps to identify the cause (and timing) of this churn opening up opportunities to implement effective retention strategies and we were responsible to find the customer churn for our client to predict and apply strategies for potential customers who might leave.

2018

Tech Mahindra
Technical Support Associate, Hinjewadi Pune

Text mining / Sentiment Analysis	Twitter data analysis to check whether the sentiment of customer is positive or not.
Description	1. Connecting to twitter using API. 2. Cleaning the text data to remove all punctuations and special characters. 3. Prepare word cloud using Word cloud function. 4. Defining a function to take a sentence as input and provide a sentiment for it. 5. Splitting the sentence into list of words 6. Iterate every word and check if it is present in the positive or negative word.

Clustering	Apply clustering on customer data and find out groups of prime, Good and average customers
Description	1. Import csv file. 2. Finding the best number of clusters based on the distortion value. 3. Defining the K-Means object for best number of clusters. 4. Running the clustering algorithm on the data set. 5. Plotting scatter values.

Classification	To find out whether a customer is Prime, good or average customer.
Description	1. Import csv file. 2. Exploring data distribution of categorical variables using bar plot. 3. Remove Useless Columns. 4. Missing values treatment. 5. Visualizing the important characteristics of a dataset using bar plot. 6. Sampling the data into two parts: Training & Testing. 7. Applying Random Forest Algorithm & find Model Accuracy using f1-score.

Multi Linear regression	To find out how many dealers are required in particular are depending on number of customer and sell in the area
Description	1. Import csv file. 2. Selected required columns as per plot object. 3. Remove Useless Columns. 4. Missing values treatment. 5. Sampling the data into two parts: Training & Testing 6. Multi linear regression algorithm. 7. Measured model accuracy with MAPE & APE.

SHALAKA DINESH SONJE

JOURNEY

OTHER INFORMATION