ICML 2018

KDD 2018

Sample Paths for Probabilistic Demand Forecasting

Our paper on how different generative models can be used for probabilistic time series forecasting.

Paper

arXiv

MQTransformer: Multi-Horizon Forecasts with Context Dependent and Feedback-Aware Attention

Our paper on how neural architectures for Forecasting can be improved and made martingales through context awareness.

Paper

NIPS 2017

Amazon

A Multi-Horizon Quantile Recurrent Forecaster

Our paper on generating multi-step quantile forecasts using Recurrent Neural Networks and WaveNet style encoders and decoders geared towards time series forecasting.

Paper

Journal of Quantitative Finance

Amazon

All Roads Lead to Quantitative Finance

Paper with Nassim Taleb on why predictions like FiveThirtyEight dont correctly account for uncertainty.

Paper Discussion with Nassim Taleb

Bloomberg Quant Seminar

University of Michigan

Accurate Prediction of Electoral Outcomes

My paper on how to properly calculate and assess election forecasts. The model was the source of the win probabilities for the 2016 US Election displayed by the Bloomberg Terminal.

Paper Slides

Bloomberg Quant Seminar

Try AGAN

A lightning talk on a simple trick to (somewhat) stabilize the training of Wasserstein Generative Adversarial Networks.

Slides

SciPy

bqplot: Interactive Visualization in the Jupyter Notebook

My talk on bqplot at SciPy 2017.

GitHub Repository Video

Courant Mathematical Finance Seminar

Deep Learning: An Overview

A lecture I gave at Courant which covers the basics of Deep Learning, along with some novel applications.

Link Slides

PyData NYC

bqplot: Interactive Visualization in the Jupyter Notebook

My talk on bqplot at PyData NYC 2017.

GitHub Repository Video

PyGotham

bqplot: Interactive Visualization in the Jupyter Notebook

My talk on bqplot at PyGotham 2017.

GitHub Repository Video

PyData Seattle

bqplot: Interactive Visualization in the Jupyter Notebook

My talk on bqplot at PyData Seattle 2017.

GitHub Repository Video

ECML 2017

Scatteract

A paper on combining Deep Learning for Computer Vision techniques with concepts from the Grammar of Graphics to automatically extract data from scatter plots.

Paper

Theoretical Machine Learning

Slides from an internal talk I gave on the basics of machine learning theory.

Slides

Morgan Stanley Seminar

An Overview of the Recovery Theorem

Slides from a talk I gave at Morgan Stanley Seminar on the Recovery Theorem.

Slides

Columbia University

McKean Stochastic Differential Equations and the Particle Method

Slides from a guest lecture I gave for Math 4079 - Non Linear Option Pricing at the Columbia University Math Department.

Slides