Access to listed projects is available on request as due to University requirements projects are not yet public. Please message me on LinkedIn for access
Ikurahodiok/MSc-Thesis [OPEN]
A repository containing my MSc Thesis Scripts.
The development of a web map application with user (mobile device, -app) and admin (computer device, -api) modes within the same app. Admin mode allows questions to be set by the admin on a map, stored in a PostgreSQL database. In user mode these questions are then retrieved from the database according to criteria (closest Q, unanswered Q's, top5 hardest Q's, etc) and the user can answer them and send their responses back to another database where their scores are stored. This app also features dashboard capabilities of retrieving and plotting performance scores which change with the use and user-performance.
Language: Javascript, HTML/CSS, PostgreSQL, Tags: web-map-application, front-end, back-end, client-side, server-side, restful-API, node, dashboard, mobile-app
A comparison of 5 learning algorithms in a supervised classification task for an agri-tech setting.
Language: Python, Tags: neural-network, multi-layer perceptron, naive-bayes-classifier, decision-trees, knn-classifier, ensemble-voting, instance-based-learning, scikit-learn
Spatio-Temporal Analysis (STA) forecasting of taxi pickup requests per road in Porto City Centre. Using statistical (ST-/ARIMA) and ML (ST-SVM) methods.
Language: R, Tags: tmap, arima-forecasting, svm-regressor, spatio-temporal-arima, spatio-temporal-autocorrelation.
Route finding exercise on the Isle of Wight. Given an input start location this program finds the shortest path to the local high point (<5km) along a road network. Plot route and travel times with ascent impedance (DEM and Naismith's-rule).
Language: Python, Tags: geojson, network, naismith-rule, rasterio, dijkstra-algorithm
Solving PiP problem via Ray Casting without libraries. OOP in creating cascading Polygon, Line, Point Geometry Classes.
Language: Python, Tags: geometry, gis, raycasting
A classical quantitative human geography/environmental criminology project using demongraphic explanatory variales (Indices of Multiple Deprivation 2019) to examine correlation with burglaries in the Greater London Area. Use of fixed-effects and mixed-effects regression models (Spatial Durbin-, Spatial Error-, Spatial Lag-, Geographically Weighted- regression models).
The novel approach in this project was comparing the difference upon the analysis by aggregating point crime data to different spatial units of analysis. Burglaries when analysed at; Point ('raw' data), Line (Road/Footpath), Polygon1 (LSOA), Polygon2 (hexagonal grids) scales.
Language: R, Tags: crime, demographics, mixed-effects models
Ikurahodiok/Probabilistic-Seizmic-Hazard-Assessments-(Risk-Damage-Cost) (module: Catastrophe Risk Modelling)
Three part project:
- Ground Shaking Module: A group project using Matlab to undertake Monte-Carlo analyses of random possible earthquake events and using ground-motion-prediction-equations to translate them to experienced shaking to a building.
- Damage Module: A solo project, when given a level of ground shaking, calculating probabilities of damage severity (by exceeding different damage states) through the HAZUS building damage framework.
- Cost Module. A solo project, given a level of ground shaking, calcuating the damage and cost to a single building through the FEMA p-58 Earthquake Hazard Framework.
Language: Python, MATLAB, Tags: CAT-modelling, probabilistic-assessment, risk-assessment, monte-carlo-analysis, natural-disasters