Naum M.

Naum M.

San Francisco, California, United States
541 followers 500+ connections

Join to view profile

About

Ambitious and passionate Machine Learning Engineer looking to expand the intersection…

Activity

Join now to see all activity

Experience

  • SambaNova Systems Graphic

    SambaNova Systems

    Palo Alto, California, United States

  • -

    Atlanta, Georgia, United States

  • -

  • -

    Atlanta, Georgia, United States

  • -

  • -

    Santa Cruz, California, United States

  • -

    Santa Cruz, California, United States

  • -

    Santa Cruz,CA

  • -

Education

  • Georgia Institute of Technology Graphic

    Georgia Institute of Technology

    -

    Specialization in Machine Learning

  • -

  • -

    Activities and Societies: Jew Crew (Jewish Club), Lowell Football, Lowell Mock Trial Team, Lowell Advanced Band, Lowell Junior Spirit Committee, Lowell Advanced Choir

Courses

  • Abstract Data Types

    CMPS101

  • Analysis of Algorithms

    CSE102

  • Applied Machine Learning

    CSE144

  • Artificial Intelligence

    CS140

  • Comparative Programming Languages

    CMPS112

  • Computer Architecture

    CE110

  • Computer Systems

    CE12

  • Computer Vision

    CS6476

  • Data Structures

    CMPS12B

  • Data Visualization

    CS6730

  • Deep Learning

    CS7643

  • Distributed Systems

    CSE138

  • Game A.I.

    CMPM146

  • Graduate Algorithms

    CS6515

  • Human Factors

    PSYCH128

  • Human and Machine Learning

    CS8803

  • Knowledge Based AI

    CS7637

  • Learning & Memory

    PSYCH129

  • Machine Learning

    CS7641

  • Natural + Artificial Intelligence

    PSYC139P

  • Networks

    CSE150

  • Operating Systems

    CS111

  • Psychology of Language

    PSYCH125

  • Research Methods

    PSYC100

  • Software Engineering

    CMPS115

  • Systems for Machine Learning

    CS8803

  • Visual & Spatial Cognition

    PSYCH120

Projects

  • SlugHub

    - Present

    Architected & managed a cross-platform mobile app which serves as a hub for the resources at UCSC. The app uses web scraping from the UCSC website to generate a dictionary of classes, which is then searched from the app to see information regarding these classes. Many features are included, such as available clubs, classes, sports, and tutoring support.

    Other creators
    See project

  • NanoNet: Quantized Neural Network

    -

    – Developed a neural network for image classification that has 75% reduction in size, 55% reduction in processing time, and - 0.62% loss in accuracy from a baseline neural network using network quantization

  • Fooling Networks: Creative uses of Deep Neural Networks

    -

    – Implemented saliency maps and layer visualizations which helps model explainability in CNNs
    – Performed adversarial attacks on a performant model using gradient ascent to incorrectly predict a class

  • FAWG: Fast Array of Wimpy GPUs

    -

    – Constructed a system to distribute inference of large DNNs amongst tiny, cheap GPUs with the SAIL lab at GT
    – Featured a model partitioning scheme, as well as efficient data transfer based on model query traffic

  • Raven’s Progressive Matrices Agent: Solving Human IQ tests using AI

    -

    – Developed an agent that is able to solve the RPM Test of Intelligence for humans with 74% accuracy
    – Exploited different ML techniques such as ensemble methods and boosting, as well as other CV tools

  • Sign Language using HMMs

    -

    -Created a system that is able to distinguish between between different signs in ASL
    -Implemented the Viterbi algorithm and Hidden Markov Models for predicting the hand sign

  • Distributed Sharded Fault-Tolerant Key-Value Store

    -

    -Collaborated with a team of engineers on a KV store program using a REST API
    -Featured improvements in asynchrony, consistency, availability, and durability

  • Infinite Axis Utility System

    -

    Developed a tool that decides which actions to take, based on the utility of each action, which is able to be applied to many games which use game state.

    See project

  • Chess Move Legality Program

    -

    Crafted a program that can determine the legality of a string of chess moves and determines the illegality of moves, if any.

    See project

Languages

  • English

    Native or bilingual proficiency

  • Russian

    Elementary proficiency

More activity by Naum

View Naum’s full profile

  • See who you know in common
  • Get introduced
  • Contact Naum directly
Join to view full profile

Other similar profiles

Explore collaborative articles

We’re unlocking community knowledge in a new way. Experts add insights directly into each article, started with the help of AI.

Explore More

Others named Naum M.

0 others named Naum M. are on LinkedIn

See others named Naum M.

Add new skills with these courses

See all courses