Hi! I am
Shehran Syed
Shehran Syed
April 2024 - Present
Collecting, cleaning, and processing data for modelling crop growth in a controlled environment.
Developing and training a machine learning model to forecast crop yield in hydroponic vertical farming set-up.
Designing a computer vision algorithm for a non-destructive approach to estimating crop yield from images.
Collaborating with other researchers to report and present findings of research to journals and conferences.
Aug - Dec 2023
Grade student work and prepare gradesheets for upper-division statistics classes.
Classes: Statistical Data Science, Mathematical Statistics, Applied Statistics for Business and Economics.
Jan 2020 - Sep 2022
Instructed students at the undergraduate level in courses within the Statistics discipline.
Courses taught: Elements of Statistics and Probability, Introduction to Statistics and Probability, Visual Programming for Business.
Coordinated a team of lecturers to facilitate effective online delivery of course content during the pandemic.
Designed a dashboard to expedite reviewing and categorizing applications for emergency financial aid during the pandemic.
Sep 2022 - Dec 2023
Jan 2016 - Aug 2019
Statistical Analysis
Data Visualization
Web Development
Adobe Software Suite
Python
R
SQL
Java
HTML/CSS
A web-based application that visualizes real-time and historical weather data for any US county. It retrieves data from OpenWeather API and presents it through an interactive dashboard using Dash App in Python. The dashboard, hosted on Render, can be accessed here.
In this project, the impact of visual stimuli on neural activity in mice's visual cortex was analyzed using data from Steinmetz et al. (2019). Findings revealed an interactive effect between stimuli and neural activity. A predictive model was built achieving 85% sensitivity and 35% specificity, suggesting further research with more data and features. A detailed write-up can be found here.
Health characteristics were used to predict age-related conditions, employing logistic regression, random forest, and XGBoost classifiers, resulting in models with 85%-95% sensitivity and specificity scores. This work demonstrates the potential of machine learning in diagnosing age-related conditions, potentially reducing the need for invasive tests.