sflemming9 Goto Github PK

cs107e.github.io

Backing repository for CS107e site. Don't commit here (automatically managed). Read articles on the site.

cs229---autumn-2017---classification-of-power-plants-from-earth-observation-data

The energy sector is the second largest consumer of water globally [1]. In an era of growing concern around water risk, stakeholders are interested in quantifying water use by thermal power plants. Methodologies to estimate this water use exist but rely on oft-unreported plant characteristics, the most important of which is cooling type. We use machine learning to predict cooling types for the entire Indian thermal power sector, testing two models reliant on different data types. We use transfer learning with Tensorflow and the MobileNets convolutional neural network (ConvNet) to achieve test accuracies of over 98% for three of the four major cooling types based on satellite image data, and a logistic regression model that achieves test accuracies of at least 78% for three cooling types based on information commonly available even when geolocations are not.

embedded_systems_starter_projects

esp8266_send_receive_udp

Handles sending and receiving arbitrary messages using UDP on a ESP8266

fizzbuzz

Basic Fizz Buzz program to become familiar with Go (for SSCP)

maze-generation

Create maze

sab_vim

vim settings

satelliteimagescraper

Satellite image scraper for CS229 project

sequential_compression

A compression and decompression scheme that replaces sequential sub-strings with the starting letter and the number of letters.