Hey! I'm Ammar Sidhu.
I am a seasoned Data & GIS Analyst adept at harnessing the power of spatial & statistical analysis to address and resolve complex business challenges with a passion for data visualizations and cartography.
About Me
Hello friends, hope you are doing well. Welcome to a snapshot of my qualifications and experience.
Qualifications:
Masters of Spatial Analysis Candidate at the Toronto Metropolitan University (Sept. 2023 to PRESENT)
GIS Analyst at CartoVista (Jan. 2023 to PRESENT)
Honours Bachelor’s of Science in Applied Statistics and Geographic Information Systems at the University of Toronto (Sept. 2018 to Dec. 2023)
Contracted GIS Analyst for Toronto Fire Services (Sept. to Dec. 2023)
Contracted Statistical Analyst at Toronto Metropolitan University for Environment Canada (Feb. 2024 to PRESENT)
Research Remote Sensing Analyst at the University of Toronto for the Toronto and Region Conservation Authority (TRCA) (May to Jun. 2023)
Data Science Intern for Data Glacier (Sept. to Dec. 2023)
Proficient in FME, ArcGIS Pro, QGIS, Python, SQL, Alteryx, Tableau, Google Earth Engine, Excel, R, Jira/Confluence, Adobe Illustrator
Technical Skills
Through my educational background in GIS and Statistics, I have acquired proficiency in a toolbox that consists of a wide array of tools for conducting geospatial and statistical analysis as well as data/geoprocessing to transform messy and unstructured data into insights.
Communication & Management
My experience as a GIS practitioner from academic studies and industry work has led me to learn how to effectively communicate and manage projects through industry-standard ecosystems and software.
Data Science Portfolio
-
Nuclear Bomb Classifier
Developed a Nuclear Bomb Deployment Classifier with 92% Accuracy on imbalanced data (4 classes) that identifies the deployment method of a nuclear bomb.
-
Employee Salary Predictor
Worked with an employee salary dataset from a recruiting company containing 1,000,000 employees across 8 credentials to predict salaries. Obtained a MSE of $355 on the test dataset with the Gradient Boosting Regressor algorithm after hyperparameter tuning with GridSearchCV.
-
Heart Disease Classification
Created a binary classification model with 87% accuracy on heart disease data to help hospitals determine if a given patient has heart disease.
-
Hamilton, ON House Price Prediction
Created a spatial regression tool that estimates house prices (MSE ~ $57070.99, R^2 ~ 0.805) to help predict house prices by census tract in the city of Hamilton, ON.
-
Bulldozer Price Predictor
Predicted bulldozer prices with Random Forest Regression on time series sales prices with an RMSLE of 0.24 and R^2 of 0.88 on validation data.
-
EDA of Cab Companies: Who to Invest In?
Conducted an exploratory data analysis and correlation analysis to identify which company in New York City is more profitable and better to invest in.
-
Bank Customer Churn Classification
Developed a Random Forest Classifier that can predict if a client will subscribe to a term deposit with 93% Accuracy.
-
EDA of Breast Cancer Statistics by Race in USA
Performed an exploratory data analysis in USA’s 2020 Breast Cancer Statistics to determine difference in breast cancer rates between black and white females.
Cartographic Portfolio
-
The Relationship Between NO₂ and NDVI in Peel Region
Collaborated with the Region of Peel to compute and map NDVI (Vegetation Index) data for Peel Region with Landsat-8 and 7 satellite imagery from 2006 to 2016. Conducted a Correlation Analysis and trained 7 regression algorithms between Nitrogen Dioxide and NDVI from 2006 to 2016. Determined the Spatial Autocorrelation of NDVI with Moran’s I of 0.97, but Pearson’s r of -0.27 implying a weak negative relationship exists between the two variables in Peel Region at the dissemination level from 2006 to 2016 .
-
Space-Time Cluster Analysis of Toronto's Homicide Data
Visualized homicide data in Toronto from the Toronto Police Service portal by conducting a 3D-Emerging Hotspot Analysis from 1990 to 2022. Mapped 3D-time clusters across the city that highlighted statistically significant cold and hotspots for homicides with up-to 99% confidence.
-
Turkey-Syria Earthquakes Live-Time Web Map
On February 6, 2023, Southern Turkey and Northern Syria were struck by earthquakes nearing magnitudes of 8.0 resulting in mass-scale destruction of cities near the epicenter of the quakes. This ultimately left millions of people displaced and well over 55,000 deaths. Consequently, the world turned their attention to Turkey and Syria to provide aid to impacted citizens. The CartoVista team has collected data on the disaster to provide awareness and insights into the earthquakes that impact Turkey.
This Story Map will visually review the history of earthquakes in Turkey, the Turkey-Syria earthquakes during February, the damages caused by the quakes as well as live updates of earthquakes in Turkey.
-
World-Religions Story Map
This Story Map will visually highlight global religion demographics and deep dive into religions in Canada with the help of recent census data by subdivisions.
From the mapping, it was found that Christianity and Islam are the world’s largest religions with an increase in agnosticism globally.
Despite Christianity being the dominant religion in Canada, it is becoming more apparent that as of 2022, Ontario and British Columbia are starting to become more heterogeneous in terms of religious beliefs.
-
Business Registries in Toronto's Ward 18
This interactive web map dashboard visualizes all the businesses in Toronto’s Ward 18 by business types. The dashboard provides an interactive experience that links the the data table to the business points on the map, which allows users to search for businesses address, and pinpoint that physical location on the map. Additionally, data tips have been added so that information is provided to the user when they hover over a business on the map.
-
2023 NBA All-Star Game Map
-
University of Toronto Rainfall Catchment 3-D Mapping by Building
The following dual panel map visualizes rainfall catch for each building in the University of Toronto St. George campus. The map on top visualizes the rain volumes as a 2D map, and the bottom map visualizes the rainfall volumes by building in 3-D using ArcGIS Pro’s 3D Analyst toolbox.
-
Clustering of Technological Development by State in USA
This geospatial statistical analysis compared hierarchical clustering and k-means clustering for mapping technological development state in USA using census 2020 data. The clustering was performed in SPSS while the mapping was done using QGIS. The maps cluster the states based on their technological and financial advancements.