laujohansson/anomalydetection_vae_lstm

Introduction

Visit my page at: https://laujohansson.github.io/

Who am I

A danish data engineer. Interested in data science. Specialized in data analysis and deep learning.

Currently working with Microsoft Azure and Azure Databricks.

Part of spetlr-org and atc-net

Contributing to open-source dataplatform python tools - atc-dataplatform and spetlr.

Contribution examples

Azure Synapse - publish to multiple evironments (dev, test and prod)

Publish your development workspace to test and production. This contributes with two (on that time) missing Synapse features:

• Azure Pipeline YAML: Automatic pipeline for publishing development -> to test and prod. Find the pipeline template here and the pipeline here.
• Initialize-SynapseParameters.ps1: A script to parameterize the Synapse paramters. Either by using individual environment specific .json files or string-replacements. Find the script here.

Project overviews

The repositories on my Github mainly contain deep learning projects.

• Natural language processing - Friends Manuscript Generator (PyTorch)
• Variational Autoencoder LSTM - Anomaly detection: Detecting road defects (PyTorch)
• Image segmentation of Lung CT's using CNN and UNET (PyTorch)
• Generative Adversarial Networks - Converting horses to zebras and generating fake handwritten digits (PyTorch)
• Convolutional Neural Networks and Sliding window - Classify hotdogs and not-hotdogs. Detecting house street numbers (PyTorch)
• Topic Modelling and FFNN - Predicting customer churn at brazillian e-commerce Olist (Keras)
• Probalistic Graph Model (Linear Dynamical System) - Predicting Covid19 (NumPyro)

Data science projects at https://laujohansson.github.io/

Me and some of my fellow students have made several data science projects.

Check it out at https://laujohansson.github.io/.

Check out the repository contains materials used on our main website here

Contact

You can reach me at my LinkedIn profile.