1. Worked on data gathering, data cleaning, data modelling,

2. Worked on explorative data analysis

3. Outlier and missing value detection and tratement

4.Data Imbalance

5. Features Selection

6. Building Classification Model

7. Evaluating Accuracy and other metrics

8. Parameter Tuning

9. Finalize the Model5.

1. Applied the hierarchical clustering and plotted the dendrogram.

2. Identified the 3 cluster- High spending, medium spending and Low spending - with help of dendrogram

3. compared the results with k-means clustering.

4. By making use of this data, company can announce various offers to various segments

5. Appliedhierarchical clustering , Agglomerative Clustering, Kmeans clustering, Python, Pnadas, numpy

Wrote data pre-processing scripts

1. Normalization: is to clean data to obtain better features
2. converting all letters to lower or upper case
3. removing numbers
4. removing punctuations
5. removing stop words, sparse terms, and particular / rare words
6. spelling correction
7. stemming: removal of suffices,
8. Lemmatization: converts the word into its root word, rather than just stripping the suffices.

2. Worked on Feature extraction by Tf-idf, CountVectorizer, Singular Vector Decomposition (SVD) and Feature Engineering
3. This is developed using Python, NLP and Machine Learning like Naïve Bays, XGboost.

Task was to predict attrition of the employees given the behavioural data and related historic attrition. Decision Tree, Random Forest

A main area offocuse is machine learning model that can identify toxicity in on line conversation where toxicity is define as some thing like disrespectful statment. This is NLP project

Gennerating a meaningful summery from the different type fo text

Built Logistic Regression and Decision Tree models to identify customers likely to default based on demographic and financial data. Evaluated performance metrics like AUC, F1 score, Recall, Precision, and Accuracy.

Identified authors of sentences in a dataset using Nave Bayes, text analysis, and feature extraction techniques like Tf-idf, CountVectorizer, and SVD.

Applied hierarchical clustering to identify customer segments based on spending behavior. Used dendrogram and k-means clustering to analyze and offer targeted promotions.

Predicted employee attrition using Decision Tree and Random Forest based on behavioral data and historic attrition patterns.