Maxim Lakin

Maxim Lakin

Mentor
5.0
(1 reviews)
US$25.00
For every 15 mins
1
Sessions/Jobs
free badge
First 15 mins free for your first session
ABOUT ME
Deep learning engineer focused on state of the art applications in pattern recognition, optimization, anomaly detection, reinforcement learn
Deep learning engineer focused on state of the art applications in pattern recognition, optimization, anomaly detection, reinforcement learning, and automation.

Hi I'm a deep learning engineer focused on statistical modeling, optimization, generative models and bayesian methods, simulations, reinforcement learning, planning and control, forecasting and predictive modeling, recommendations, and anomaly detection

Here are some projects I've done in the past:

  • Built an A3C reinforcement learning model to optimize gas production on multi-well pad using time-series, computer vision, and a physics-based simulator for training, testing, and validation.

  • Developed a computer vision and reinforcement pipeline for an battle bot competition between fully autonomous and human controlled bots using Unity Agents, Tensorflow, OpenCV

  • Developed hierarchical scene understanding recognition model capable of tracking variable number of objects in multiple configurations

  • Built a multi-armed bandit model for online control bootstrapping exploration with historical data to control heating in a system with difficult to simulate high sensitivity to initial conditions

English
Central Time (US & Canada) (-05:00)
Joined November 2014
EXPERTISE
4 years experience
10 years experience
5 years experience
4 years experience
10 years experience
5 years experience

REVIEWS FROM CLIENTS

5.0
(1 reviews)
Nathan R
Nathan R
December 2014
Max was very knowledgeable and helpful. He worked with me through my project and helped me solve my problem. I highly recommend him!
EMPLOYMENTS
Deep Learning Engineer
Markov
2018-10-01-Present
Built a multi-armed bandit model for online control bootstrapping exploration by using historical nearest neighbor lookup to control heat...
Built a multi-armed bandit model for online control bootstrapping exploration by using historical nearest neighbor lookup to control heat diffusion in a system with difficult to simulate high sensitivity to initial conditions Developed hierarchical scene understanding recognition model capable of tracking variable number of objects in multiple configurations Used variational autoencoder to learn latent dynamics, propose policies, and explore outlier phenomena
Python
OpenCV
TensorFlow
View more
Python
OpenCV
TensorFlow
View more
Deep Learning Engineer
Kelvin Inc.
2017-06-01-2018-10-01
Built an A3C reinforcement learning model to optimize gas production on multi-well pad using time-series, computer vision, and a physics-...
Built an A3C reinforcement learning model to optimize gas production on multi-well pad using time-series, computer vision, and a physics-based simulator for training, testing, and validation. Developed a convolutional variational autoencoder for unsupervised feature extraction for simulations. Developed an LSTM classifier and anomaly detection pipeline for reporting and performance evaluation
Python
NumPy
Pandas
View more
Python
NumPy
Pandas
TensorFlow
View more
Deep Learning Engineer
AIorDie
2017-06-01-2018-09-01
Developed a computer vision and reinforcement pipeline for an battle bot competition between fully autonomous and human controlled bots u...
Developed a computer vision and reinforcement pipeline for an battle bot competition between fully autonomous and human controlled bots using Unity Agents, Tensorflow, OpenCV Implemented hardware and sensor architectures and deployed models to bots running Odroid and Mbed, built telemetry and policy visualizations
Python
C++
OpenCV
View more
Python
C++
OpenCV
TensorFlow
View more