firefly - MScR Astrophysics

A target selector for use with TransitFit to fit TESS lightcurves.

---

serenity

A gradient descent ensemble boosted tree (regressor!) to compare predictions versus Forecaster.

Predicting the Mass:

+-------------+---------+--------------+--------------+----------------------------+----------------------------+---------------------------+
| Planet      |    Test |   Prediction |   Forecaster |   Residual Test-Prediction |   Residual Test-Forecaster | Prediction < Forecaster   |
|-------------+---------+--------------+--------------+----------------------------+----------------------------+---------------------------|
| Kepler-17 b | 2.47000 |      2.17690 |      3.70134 |                    0.29310 |                    1.23134 | True                      |
| Kepler-56 b | 0.74300 |      0.11561 |      0.11645 |                    0.62739 |                    0.62655 | False                     |
| XO-1 b      | 0.91300 |      0.75850 |      5.03887 |                    0.15450 |                    4.12587 | True                      |
| WASP-72 b   | 1.54610 |      2.13558 |      2.43534 |                    0.58948 |                    0.88924 | True                      |
| HATS-6 b    | 0.31900 |      1.00091 |      7.65828 |                    0.68191 |                    7.33928 | True                      |
| WASP-7 b    | 1.25000 |      1.65510 |      9.44126 |                    0.40510 |                    8.19126 | True                      |
| Kepler-29 c | 0.01259 |      0.05699 |      0.03317 |                    0.04440 |                    0.02058 | False                     |
| L 98-59 c   | 0.00761 |      0.00977 |      0.00766 |                    0.00216 |                    0.00005 | False                     |
| TrES-2 b    | 1.19800 |      1.14120 |      6.21201 |                    0.05680 |                    5.01401 | True                      |
| WASP-32 b   | 2.63000 |      1.89571 |      9.44126 |                    0.73429 |                    6.81126 | True                      |
| KPS-1 b     | 1.09000 |      1.46330 |      7.65828 |                    0.37330 |                    6.56828 | True                      |
| K2-266 e    | 0.04499 |      0.02528 |      0.02183 |                    0.01971 |                    0.02316 | True                      |
| HAT-P-41 b  | 0.79500 |      1.09818 |     30.01569 |                    0.30318 |                   29.22069 | True                      |
| KELT-7 b    | 1.28000 |      2.50883 |      4.08728 |                    1.22883 |                    2.80728 | True                      |
| HAT-P-13 b  | 0.90600 |      0.90435 |      3.00234 |                    0.00165 |                    2.09634 | True                      |
+-------------+---------+--------------+--------------+----------------------------+----------------------------+---------------------------+

Residual sum for Prediction: 5.52
Residual sum for Forecaster: 74.97
Prediction versus Forecaster Accuracy: 12/15

---

Radio Interferometer Simulation - MScR Astrophysics

A simulation of the VLA Radio Interferometer.

---

dataSciencePortfolio

A collection of projects and notebooks as examples of my work, adopting commonly used machine learning algorithms.

• Amazon Food Reviews - A collection of Kaggle projects to self teach many various ML/AI algorithms.
• Minimax Algorithm - Noughts and Crosses
• Ray Tracing - Linear Interpolation in 3D using numpy broadcasting instead of for loops.
• Gaussian Noise Galaxy Simulation - A galaxy simulator created using gaussian noise.
• Decision Trees - Detecting Breast Cancer
• Linear Regression - Boston House Prices
• Neural Network - Prototyping
• OOP Simulation - Coffee Shop

---

Certificates

Intro to Machine Learning

• Data Exploration
• Model validation
• Underfitting and Overfitting
• Random Forests

Intermediate Machine Learning

• Handle Missing Values
• Categorical Variables
• Pipelines
• Cross-Validation
• XGBoost
• Data Leakage

Machine Learning Explainability

• Use cases for model insights
• Permutation Importance
• Partial Plots
• SHAP Values
• Advanced Uses of SHAP Values

---