MLH's Top 50 Hackers

Most inspirational hackers out of 100,000 hackers around the globe.

Aria came to the United States by herself as an international student at the age of fifteen. She had to teach herself English and is proud to have completed her education on her own in another country and secured her dream job as a software engineer at MediaMath. Aria’s exposure to the hacker community stemmed from a fascination in the math behind artificial intelligence, machine learning, and game AI algorithms. However, Aria lacked confidence as she had little experience in coding prior to her sophomore year of college. When she finally decided to take more computer science courses during her junior year, she felt behind compared to her peers. Soon, Aria found that she had nothing to worry about: she won two portfolio at her first hackathon at UC Berkeley. Aria’s team worked with Google Cloud’s Speech to Text API and IBM’s Watson’s Sentiment Analysis API. It was a lot of work to get the project to track student’s learning progress, but it was worth it. Their project won the best education hack and CalHacks fellowship award. This gave Aria the validation and recognition she needed to continue pursuing computer science. Addicted to the high of winning, Aria started attending hackathons left and right. Hackathons inspired Aria to jump into projects in unfamiliar domains. She has taken up multiple research projects with professors and created her own side projects. Hackathons also led her to an internship doing smart grid research. In this internship, she created the prototype for a deep reinforcement learning agent to control power grid settings and fix contingencies. The paper based on the model she created won her colleagues and her the best IEEE paper award. She received an offer from MediaMath for a full-time software engineering position on a data engineering team doing large scale machine learning projects. Her favorite project came from her team at HackDavis when she worked with the Muse headband which utilizes brainwave detection. Their project Brainalyzer used the headband to record a user’s brain waves in order to detect abnormalities and predict seizures. They found a brainwave dataset and performed feature selection and created an LSTM model to predict brainwaves. This project won the best use of Google Cloud Platform.

• Date: February 2020
• Award: MLH'S TOP 50 Most Inspirational Hackers
• Organization: Major League Hacking

Close Project

Grid Mind

Autonomous Voltage Control for Grid Operation Using Deep Reinforcement Learning

Abstract—Modern power grids are experiencing grand challenges caused by the stochastic and dynamic nature of growing renewable energy and demand response. Traditional theoretical assumptions and operational rules may be violated, which are difficult to be adapted by existing control systems due to the lack of computational power and accurate grid models for use in real time, leading to growing concerns in the secure and economic operation of the power grid. Existing operational control actions are typically determined offline, which are less optimized. This paper presents a novel paradigm, Grid Mind, for autonomous grid operational controls using deep reinforcement learning. The proposed AI agent for voltage control can learn its control policy through interactions with massive offline simulations, and adapts its behavior to new changes including not only load/generation variations but also topological changes. A properly trained agent is tested on the IEEE 14-bus system with tens of thousands of scenarios, and promising performance is demonstrated in applying autonomous voltage controls for secure grid operation.

• Date: August 2019
• Status: In production
• Category: IEEE Power and Energy Society General Meeting Best Conference Paper Award

Close Project

Brainalyzer - Seizure Predictor

HackDavis 2019

The application targets patients with brain abnormalities such as seizure and epilepsy. We analyzed the EEG waves as a time series prediction task using Machine Learning. The pre-trained model predicts the future occurrence of seizure by analyzing the EEG brainwaves of the user. By constantly wearing the Muse Headband, the user feeds input to the model to improve its prediction accuracy. With optimal hyperparameter selections, our model can predict new patients' wave patterns AND whether or not a seizure is likely to occur at a future timestamp with 91% accuracy.

• Date: February 2019
• Award: Google Cloud Platform
• Awarding Organization: Major League Hacking

Close Project

AEye

Skin disease detection app.

Our ML model has been trained on hundreds of samples of diseased skin to be able to identify among a wide variety of malignant and benign skin diseases. We have a mobile app that lets you take a picture of a patch of skin that concerns you and runs it through our model and tells you what our model classified your picture as. Finally, the picture also gets sent to a doctor with our model results and allows the doctor to override that decision. This new classification is then rerun through our model to reinforce the correct outputs and penalize wrong outputs, ie. adding a reinforcement learning component to our model as well.

• Date: January 2017
• Client: Lines
• Category: Branding

Close Project

E[Leap]ments Leap Motion AI Game

An interactive puzzle game which challenges you to defeat an AI opponent in a captivating Leap Motion interface.

Our project is an interactive game, similar in intent to the universally-loved game Codebreaker, now enhanced by an artificially intelligent opponent and presented in an appealing interface that introduces the user to the new capabilities of Leap Motion in a comfortable, 8-bit-inspired UI. The user and the AI-opponent face off in a “player-vs-player” viewpoint. The user has access to four hand gestures, each mapping to one of the four elements: Air, Fire, Earth, and Wind. Each has one goal: discover the most powerful combination of these elements (in other words: a 4-action combination) in order to master the elements before their opponent does and defeat them. While the user becomes more comfortable with 4-action combinations, the opponent learns along with them, and responds to an increase in player skill by also becoming more skilled at the game. This creates an incentive for the player to seek new and more powerful element combinations- the best of which randomize at the beginning of each game- to win the game. In addition, sounds for this game were all original. They were created using an electronic music editor, FL Studio.

• Date: January 2017
• Client: Southwest
• Category: Website Design

Close Project

Best Education Hack + Fellowship Awards

My first hackathon @calhacks2018

Smarter Feedback for Webcasts. Our Web Application allows students to better interact with their teachers by marking time-intervals where they feel confused in the webcast. With this information, we provide teachers feedback on sections of lecture and topics they may want to revisit. In the future, we plan to build more sophisticated feedback systems to give teachers more access to information. We will include comments along with the timestamps to let students have specific queries along with the ability to mark their level of confusion. Our ultimate hope for this is to create a more immersive and personal webcast experience by allowing students to give more input in an non-obstrusive, streamlined manner.

• Date: November 2018
• Awarding : CalHacks at UC Berkeley
• Organization: Major League Hacking

Close Project