Maryam Amjad

I am a passionate data scientist, language enthusiast, and relationship builder with a background in computer science. I combine my technical skills in math, statistic, and analytics with my transferable skills in time management and teamwork to contribute to an organization’s success.


Experience

DATA SCIENCE IMMERSIVE

GENERAL ASSEMBLY, NY
  • Acquired, cleaned, and explored large datasets using Python, SQL, and Tableau to present findings to both technical and non-technical audiences.
  • Completed 40 labs and 5 projects focused on real-world applications of data science principles and best practices.
  • Developed a Natural Language Processing model to classify mobile OS using subreddit posts.
  • Worked on a team to use Machine Learning to make predictions and determine the outcome type of animals in Austin Animal Center.
June 2021 - Sep 2021

Tech intern

NYC Department of Education, NY
  • Troubleshoot/ corrected WIFI errors.
  • Reviewed steps to identify source at error.
  • Inventory- record all the data from school about the computers which don’t work.
  • Explained steps/ guided students with different apps and procedures.
  • Ensure that everything related to technology is working fine
May 2019 - March 2020

Education

GENERAL ASSEMBLY, Manhattan, NY

DATA SCIENCE IMMERSIVE
June 2021 - Sep 2021

Hunter College, Manhattan, NY

BA - Computer Science
2019 - 2021

Data Science Projects

Credit Card Fraud Detection

Credit card fraud costs consumers and the financial institutions billions of dollars annually, and fraudsters continuously try to find new tactics to commit illegal actions. With this level of control, fraudsters don't have the chance to make multiple transactions on a stolen or counterfeit card before the cardholder is aware of the fraudulent activity. This alone can save a significant amount of money that would normally be lost to fraud.

Animal Outcomes Prediction - Austin Shelters

Austin Animal Center has different kinds of animals that come in and go out based on different aspects. The staff are facing issues in facilitating the shelters for the animals as they are unsure of the outcome type of these animals. So it is decided to predict theoutcome type of the animals, which is vety much important for facilitation. The data from the past is cleaned and analysed. Ultimately, the project team was successful in crafting a model that performs well above the baseline when predicting whether or not an animal is adopted. Key variables include duration of stay and animal age. As the duration of stay increases, so too does likelihood of adoption. As age increases, adoption likelihood decreases.

My contribution to this project includes data cleaning and EDA. In the cleaning phase, any inconsistency in data is identified and removed. In EDA phase the relation between the features are analysed and suitable predictors are identified.
Know more

Binary Classification with NLP

Web Scraping is an automatic method to obtain large amounts of data from websites. Most of this data is unstructured data in an HTML format which is then converted into structured data in a spreadsheet or a database so that it can be used in various applications. Reddit has an API called Pushshift API that allows you to access their data in a structured format. This API is then used to scrap the data which is then cleaned and analysed. NLP technique is used and the contents are classified.

Skills

Data Science
  • Python
  • Pandas
  • Numpy
  • Sci-kit Learn
  • Matplotlib
  • Natural Language Processing
  • Neural Network
  • Data Cleaning
  • Exploratory Data Analysis
  • Visualizations
  • Modeling
  • SQL
Microsoft Office Suite
  • Word
  • PowerPoint
  • Excel
  • Outlook
  • OneNote