•  Worked on multiple projects during this internship:

 1) Topic Modeling: 
      This project aims to find the set of topics that best describe a document.

      • Data: used the ABC NEWS dataset, it contains data of news headlines 
          published over a period of 14 years.
      • Algorithms: 
                   •  Stemming and Lemmatization as a text preprocessing step.
                   •  TF-IDF.
                   •  LDA  for classifying Documents.

2)  Movie Rating Prediction:
      This project aims to predict the Rating of a movie based on a set of attributes.

      • Data : used the Netflix Movies and TV Shows This  dataset consists of listings of all 
            the movies and tv shows available on Netflix, long with details such as - cast, directors, ratings, release year, duration, etc.
      • Algorithms : 
                    •  Data Cleaning, Feature engineering and Preprocessing.
                    •  Logistic Regression & Linear Regression.

3) House Price Prediction competition: 
      This project aims to predict Price of a house based on a set of attributes.

      • Data : used the Housing Prices with 79 explanatory variables describing (almost) every aspect of 
              residential homes in Ames, Iowa, this competition challenges you to predict the final price of each home.
      • Algorithms & Code : 
                  •  Data Cleaning, Feature engineering and Preprocessing.
                  •  XGBoost. 

        Me :)
        
        

                            

This internship was focused to help solve a large data problem that faced the company,
Problem Description:
  the company had more than 100k records of companies that had a  lot of attributes for each company 
  including name,domain,contacts, etc.
  but those records are duplicated in a fuzzy way, in which there are multiple names for the same company 
  because of data entry and old company policy issues.
  Hence, the task was to reduce the number of duplicated records based on name similarity and data analysis process.   

Solution:
    • Technologies:
      ⚬ Python
      ⚬ Sql

    • Approach:
      ⚬ first i used fuzzy matching algorithms within the fuzzy-wuzzy library to find companies based
        on their name similarity using a python script.

      ⚬ then i went deeper into data analysis using Pandas library to find another attributes which can contribute to finding
        duplicate companies, like: domain name, region, email correspondence, etc.
      
      ⚬ after that i classified companies into:
               ▬  Deleted : company records that are not necessary.
               ▬  Merged : company records that need to be merged with their fuzzy duplicate.
              then we made stored procedures in sql to handle these companies.
               
      ⚬  By using pyodbc library i was able to connect to the company's live sql server, and with a python
                script i was able to solve the company's data problems and retain a better and more valuable database for the comapany.