PyData Amsterdam 2017
2017-04-07..2017-04-09
Amsterdam, Netherlands
[PyData Amsterdam 2016, PyData Amsterdam 2017, PyData Amsterdam 2018, PyData Amsterdam 2019, PyData Amsterdam 2024, PyData Amsterdam 2025]
| title | author | length |
|---|---|---|
| Deep Learning for Time Series Made Easy | Dafne van Kuppevelt | 23m |
Shows an actual tool and workflow rather than vague promises: an open source library mcfly that acts as a wrapper around Keras and runs a random hyperparameter… | ||
| Neural Networks for Recommender Systems | Maciej Kula | 33m |
Neural recommender prototypes use embedding layers in PyTorch with dot‑product scoring, Siamese pairwise losses and an adaptive negative sampler. Code and… | ||
| Deploying Python models to production | Niels Zeilemaker | 32m |
Deploys Pandas/Scikit models as Docker containers built by Jenkins pipelines, pushed to a registry and launched on Kubernetes with liveness/readiness probes… | ||
| Bayesian optimization with Scikit-Optimize | Gilles Louppe | 29m |
PyData Amsterdam 2017
You are given access to an espresso machine with many buttons and knobs to tweak. Your task is to brew the best cup of espresso… | ||
| Smoothing Your Data with Polynomial Fitting: A Signal Processing Perspective | Cees Taal | 26m |
The talk explains Savitzky–Golay polynomial smoothing and analyzes it from a frequency-domain / signal-processing viewpoint, showing that polynomial fitting… | ||
| Deep Learning at Booking.com | Emrah Tasli, Stas Girkin | 33m |
Booking.com adapts convolutional nets to its hotel-image corpus and runs 200,000 images per day through custom multilabel classifiers, links image labels to… | ||
| A Pythonic Tour of Neo4j and the Cypher Query Language | Nigel Small | 22m |
Overview claims how to use Neo4j from Python with concrete tools and demos: the neo4j-driver (Python) using the Bolt protocol, Cypher queries parameterized for… | ||
| Deep Reinforcement Learning: theory, intuition, code | Maxim Lapan | 28m |
Claims concrete implementations of DQN PG and A3C less than 200 lines of python using Keras and shares a code link; demonstrates Atari gameplay with a… | ||
| Applied Data Science | Giovanni Lanzani | 35m |
Claims that production fails for lack of engineering detail not theory and shows concrete cases: a Spark pipeline goes from 2 to 20 hours when a random forest… | ||
| Survival analysis for conversion rates | Tristan Boudreault | 22m |
Shopify reframes conversion as time‑dependent and shows that a snapshot 5.8% (800 of 14,000) after one week underestimates eventual spenders; cohorts diverge… | ||
| A billion stars in the Jupyter Notebook | Maarten Breddels | 31m |
Vaex exposes columnar HDF5-backed tables and processes a billion rows per second by memory‑mapping and streaming chunked expressions to compute N‑dimensional… | ||
| Successfully Applying Bayesian Statistics to A/B Testing in Your Business | Ruben Mak | 39m |
The talk examines using Bayesian statistics—especially hierarchical Bayesian models implemented with PyMC—to improve A/B testing in business. It compares… | ||
| Diagnosing Machine Learning Models | Lucas Javier Bernardi | 39m |
Linear OLS on 17,000 rows and ~40 inputs shows huge coefficient variance and many nonsignificant p values until one-hot encoding and redundant month/season… | ||
| Siamese LSTM in Keras: Learning Character-Based Phrase | Carsten van Weelden, Beata Nyari | 30m |
Claims a working pipeline that maps noisy job-title strings into a 5,000+ class taxonomy by learning character embeddings with a B-LSTM encoder trained as a… | ||
| Training a TensorFlow model to detect lung nodules on CT scans | Mark Jan Harte | 26m |
A Python TensorFlow pipeline trains a two-stage model that first generates voxel-wise candidate maps on 200×512×512 CT volumes (dataset ~5 TB) and then… | ||
| A practical guide to speed up your application with Asyncio | Niels Denissen | 34m |
Shows a working Marktplaats scraper that switches from requests + synchronous writes to asyncio with aiohttp and asyncpg, creates ~100 concurrent tasks (one… | ||
| Using deep learning in natural language processing | Rob Romijnders | 26m |
Describes Google Translate moving from phrase‑based to a seq2seq recurrent neural network with attention trained across about 26 languages, using data and… | ||
| Debugging PySpark - Pretending to make sense of JVM stack traces | Holden Karau | 40m |
Covers concrete debugging tactics for PySpark: where to find useful output (use the spark history server or worker stderr rather than the console), how to… | ||
| Data Science in the Internet of Things with Python and Spark | Rafael Schultze-Kraft | 32m |
Prototype smart office collects 200+ sensors over 36 rooms for ~3.5 months via MQTT → Kinesis Firehose → S3, sensors emit on change so timestamps are… | ||
| Keynote: Ethical Machine Learning: Creating Fair Models in an Unjust World | Katharine Jarmul | 37m |
Claims that real systems reproduce societal bias with concrete examples and fixes: ProPublica COMPAS finding shows risk scores correlate with race; Google… | ||
| Finding Needles in a Growing Haystack | Stephen Helms | 31m |
An architecture that turns terabytes of trading logs and ~100k candidate time series into daily alerts using online Bayesian updating and simple conjugate… | ||
| Pythonic Metal | Iain Barr | 27m |
Counts on a scraped darklyrics corpus of ~200k songs (~110k English) produce concrete results: a metalness (log ratio) score ranks “Burn” as the top metal… | ||
| Risk Analysis | Rogier van der Geer | 31m |
A Python simulation of Risk plus a genetic algorithm produces parameterized bots whose weights evolve using TrueSkill ranking over hundreds of parallel… | ||
| xtensor: The Lazy Tensor Algebra Library | Sylvain Corlay, Johan Mabille | 20m |
xtensor is a C++ template tensor algebra library that implements lazy broadcasting and universal functions with a numpy-style API and STL idioms, claiming… | ||
| Creativity and AI: Deep Neural Nets "Going Wild" | Roelof Pieters | 34m |
PyData Amsterdam 2017
We live in times, where science fiction authors are struggling to keep up with reality. This talk will make sense of the explosion of… | ||
| Keynote: Python vs Orangutan | Dirk Gorissen | 36m |
A drone system detects implanted radio pings to locate six orangutans across ~2000 km2 by flying lawn‑mower patterns, collecting samples with a… | ||
| Simulate your language. ish. | John Paton | 28m |
PyData Amsterdam 2017
John will present a simple character-level Markov model for simulating language in Python. The goal is to generate text that… | ||
| Different Strategies of Scaling H2O Machine Learning on Apache Spark | Jakub Hava | 32m |
PyData Amsterdam 2017
This talk gives a basic overview of machine learning on top of H2O and Spark and explains different ways how to scale your tasks on top… | ||
| Lightning Talks & Closing Remarks | Ilya Tookoff | 01h 09m |
Lots of short demos and concrete tools: a warning about Python 2 end of life, a notebook testing demo using nbval pytest plugin, a repo for connecting… | ||
| nothing matches | ||