Troy_Portfolio

Data Science / Image Analysis / GIS portfolios

Project 1: Lunar Geology for Human Exploration: Project Overview

• Using ENVI and MATLAB to evaluate lunar spectral signitures for identifying selected minerals

Code and Resources Used

• ENVI: 5.3 MATLAB 6.5
• Packages/Libraries: MMM, Lunar Spectral Library, JMUSTARD, TLROUSH, RVMORRIS, ISAACSON, EXCEL, DEM
• Datasize >100 GB

Keywords: planetary geology, PSR, water ice, artemis, space exploration, ENVI, Matlab, spectral signitures

Project 2: Data Science Purchase Estimator: Project Overview

• Created a tool deployed on AWS Sagemaker that predicts the likelyhood of bank customers making a purchase
• Using XG-BoostML_Algorithms to train, test and predict employing a confusion matrix.

Code and Resources Used

• Python Version: 3.7
• Packages/Libraries: boto3, re, sys, math, json, os, sagemaker, urllib.request, numpy, pandas, matplotlib, pyplot
• Dataframe Shape (28831, 61) (12357, 61)

Project 3: Data Science IOT Anomaly, Risk, and Failure Prediction Detection: Project Overview

• Cleaned datasets can be implemented from IoT devices
• We format your data for time series analysis.
• You will gain;
• How your sensors perform,
• We find anomalies which can identify IoT sensors at risk.
• Python code which can be run on AWS Sagemaker or a desktop Jupyter notebook

Code and Resources Used

• Python Version: 3.7
• Packages/Libraries: pandas, numpy, matplotlib, pyplot, seaborn, sklearn, LabelEncoder, train_test_split, LogisticRegression, confusion_matrix, classification_report, accuracy_score
• Dataframe Shape: (220320, 52)
• Additional References
• Techniques to Handle Imbalanced Data
• Failure of Accuracy

Project 4: Data Science Python, Keras A.I., TensorFlow, and SQL tool to analyze, manipulate, and predict missing data: Project Overview

• Any dataset of three or more time series can be implemented to predict the gaps in one of the datasets
• This project can help you to learn:
• how to analyze, manipulate, and predict missing data,
• how to label, plot, and visualize data,
• how to detect correlations,
• how to transform or scale data,
• how to use Keras A.I. libraries for training data and predicting missing observations,
• python code which can be run on AWS Sagemaker or a desktop Jupyter notebook

Code and Resources Used

• Python Version: 3.0
• Packages/Libraries: pandas, numpy, matplotlib, datetime, seaborn, sklearn, StandardScaler, keras, Sequential, and Dense
• Dataframe Shape: (4, 724) to any large scale
• Additional References : None Yet