PyData San Francisco 2016
2016-08-12..2016-08-14
San Francisco, United States
| title | author | length |
|---|---|---|
| Searchable datasets in Python | Dani Ushizima, Flavio Araujo, Romuere Silva | 35m |
A Python package called pyCBIR performs content‑based image retrieval across domains, exporting signatures to CSV and offering GUI and CLI. It implements six… | ||
| Applying machine learning to software development to reduce bugs | Nitin Borwankar | 38m |
Claims that machine learning can locate concentrated risk in codebases and produce file-level risk scores using example Apache data. Demo builds a data… | ||
| Architectures for Exploring Your Personal Data | Candida Haynes | 40m |
Project assembles social exports: a Twitter ZIP with 4,000 tweets as CSV and HTML, LinkedIn archives with click‑through logs, and Google bundles. Workflow uses… | ||
| TrailDB tutorial Store and process billions of events efficiently | Ville Tuulos | 01h 26m |
TrailDB claims an immutable C library with Python bindings that stores billions of events and compresses datasets tightly — e.g. a Wikipedia dump of ~600… | ||
| Pandas, Data Wrangling & Data Science | Krishna Sankar | 01h 31m |
Hands-on Pandas usage with notebooks at https://github.com/xsankar/cautious-octo-waffle shows exact steps: read CSVs, parse timestamps, use groupby to get… | ||
| Build Data Apps by Deploying ML Models as API Services | Ramesh Sampath | 01h 18m |
Demonstration converts a trained scikit-learn KNN on iris into a pickled artifact, wraps it in a Tornado async API that uses a thread pool to run blocking… | ||
| Introduction to Julia | Tony Kelman | 02h 00m |
Julia claims to remove the two‑language problem by using LLVM JIT so high‑level code runs like C and shows that by implementing Newton’s method, calling C… | ||
| Data Applications with Bokeh | Bryan Van de Ven | 02h 14m |
Claims Bokeh lets Python drive rich browser apps and proves it: live Gapminder-like plots without a server, full interactive apps via a Bokeh server, and a… | ||
| Keynote: Empowering people by democratizing data skills | Tracy Teal | 43m |
Data Carpentry and Software Carpentry run two-day workshops that teach concrete skills — shell scripting on Amazon cloud instances, GitHub version control… | ||
| Building domain specific databases using Python for prototypes during take off | Josh Yudaken | 33m |
Prototype implements a Sliding HyperLogLog unique counter using Kafka & Python as a durable stream and in-memory prototype, then ports to RocksDB and C++ with… | ||
| Community sustainability in Wikipedia a review of research and initiatives | Stuart Geiger | 33m |
Review of Wikipedia research and initiatives cites concrete findings and interventions: editor retention drops below 10% by 2011, automated moderation via… | ||
| Time Series for Python with PyFlux | Ross Taylor | 42m |
PyFlux is a Python library that implements a broad set of time‑series models and inference methods: PyFlux, score-driven (GAS) models and variational… | ||
| Nipype | Chris Gorgolewski | 37m |
Nipype unifies hundreds of brain imaging tools into Python wrappers and actually runs command‑line, MATLAB compiled and native tools, producing reproducible… | ||
| The "NumPy" Approach | James Powell | 35m |
Shows a concrete performance claim: rewriting a binary genetic algorithm from Python lists and explicit loops into NumPy arrays and vectorized ops cuts runtime… | ||
| JupyterLab | Jamie Whitacre, Brian E. Granger | 37m |
JupyterLab ships as an early alpha that aims to replace the Notebook while keeping the same server; adoption estimates run between 1–3 million users and over… | ||
| Keynote: Python for Pythonistas | Peter Wang | 38m |
PyData SF 2016 marks the 20th meeting and roughly 230 talks over four years while Python adoption accelerates into deep learning, scientific computing and… | ||
| Fighting Against Chaotically Separated Values with Embulk | Sadayuki Furuhashi | 42m |
Claims that Embulk auto-detects chaotic CSV dialects, runs as an embeddable Java library and a command-line tool, and supports 100+ community plugins for… | ||
| Moving Forward Through The Darkness | Chia Chi Chang | 40m |
States specific modeling blind spots and how they are fixed: data–problem mismatch that makes post logs useless for purchase prediction, loss of signal during… | ||
| Large Scale CTR Prediction Lessons Learned | Florian Hartl | 31m |
Yelp runs logistic regression with thousands of features and millions of samples and reports that model training on Spark via MRJob/AWS EMR completes on the… | ||
| Resampling techniques and other strategies | Ajinkya More | 44m |
Dataset is synthetic from scikit learn make_classification and experiments use logistic regression plus the imbalanced‑learn library. Concrete claims… | ||
| LIghtning Talks | 01h 05m | |
Quick, specific signal: demos and numbers matter. See a live Keras demo that learns y = x^2 in a few lines and compares TensorFlow code size; watch a data… | ||
| Building Recommender Systems Using Python | Divya Sardana | 01h 38m |
Hands-on walkthrough that actually codes three implementations in Python — popularity, classification-style, and item‑based collaborative filtering — using… | ||
| What's New in Python 3 | James Powell | 01h 22m |
Slides list concrete Python 3.5–3.6 changes: PEP-based survey finds ~42 relevant PEPs, shows matrix @ operator implemented and used by numpy, general unpacking… | ||
| Python Visualization for Exploration of Data | Stephen F Elston, Ronald Lopez | 01h 28m |
Hands-on tutorial that actually shows code, notebooks and a GitHub repo using matplotlib, pandas plotting, seaborn to explore a small 1985 auto dataset (about… | ||
| Applied Time Series Econometrics in Python and R | Jeffrey Yau | 01h 40m |
Two-hour hands-on tutorial that derives ARIMAX and SARIMAX equations, shows ACF/PACF identification rules, and runs estimation and diagnostics in Python using… | ||
| Keynote: Using the Python Data Science Stack to Determine | Megan Price | 38m |
Human Rights Data Analysis Group processes four source lists into a deduplicated roster using record linkage then plans a separate multiple systems estimation… | ||
| Overview of the NumFOCUS Data Science Stack | James Powell | 01h 19m |
A concrete inventory of what NumFOCUS actually funds and why: lists of Jupiter notebook and numpy and libraries (pandas, matplotlib, scikit-learn, xray… | ||
| Anaconda Ecosystem for Open Data Science | Ian Stokes Rees | 36m |
Anaconda claims to bundle over 200 tools and libraries with quarterly releases and uses Conda to build isolated, link‑based environments that the presenter… | ||
| Clustering Data Science Interviews Seven Related but Distinct Categories | Samantha Zeitlin | 37m |
Claims seven distinct interview types based on more than 30 phone screens and a short taxonomy that maps roles to the data stack: analyst (SQL, Excel)… | ||
| Carousel A Python Model Simulation Framework | Mark Mikofski | 44m |
Claims an extensible modeling framework that extracts boilerplate (data I/O, unit conversion, reporting, uncertainty, simulation control) so teams can focus on… | ||
| Virtual Laboratories Vision for an Earth Sciences | Charles Doutriaux | 31m |
Proposal merges Earth System Grid Federation and Climate Data Analysis Tools into Virtual Laboratories that bring code to data on distributed storage, not by… | ||
| Visualizing Geographic Data With Python | Christopher Roach | 31m |
Explains why projections matter (Greenland vs Africa, Mercator incorrect and unfair but useful), shows concrete code paths: build a Robinson world map and a… | ||
| Altair Declarative, Statistical Visualization for Python | Brian E. Granger | 42m |
A Python library that implements a constrained declarative grammar for tidy, DataFrame‑based plots, guaranteed to emit Vega‑Lite JSON rather than render… | ||
| Creating Knowledgebases from unstructured text | Sanghamitra Deb | 37m |
Claims that you can skip large labeled corpora by encoding domain heuristics into labeling rules and letting snorkel turn those rules plus ~100 human labels… | ||
| A hybrid approach to model randomness and fuzziness | Taposh Roy, Austin Powell | 39m |
Prototype integrates Dempster Shafer evidence fusion with logistic regression priors to convert point probabilities into belief intervals, explicitly modeling… | ||
| How you really get your data science models into production the cool way! | Dat Tran | 33m |
Builds an API-first microservices pipeline that actually serves models: a Keras MLP on the MNIST dataset reaches ~98% accuracy, model artifacts go into Redis… | ||
| Machine Learning vs The Flu | Rohan Koodli | 28m |
Predicts next-season HA and NA sequences from IRDB data, encoding A/T/G/C as numerics in 15-base chunks, and compares decision trees, extra trees… | ||
| How Soon is Now: extracting publication dates with machine learning | Julie Lavoie | 31m |
A parser implements a research paper with supervised learning to extract publication dates across sites, training on human-labeled HTML snippets and features… | ||
| Transfer Learning and Finetuning Deep Convolution Neural Network | Anusua Trivedi | 52m |
Claims reuse of an ImageNet-pretrained GoogLeNet by fixing the first two convolutional layers and fine‑tuning upper layers to transfer features to domain… | ||
| A Practical Introduction to Airflow | Matt Davis | 46m |
Airflow is presented with concrete details: a production pipeline at Clover runs about 450 tasks and uses a Postgres metadata DB, S3 for logs, and the Celery… | ||
| Closing the Machine Learning Loop using Python | Randall Shane | 50m |
A production pipeline written in Python + scikit-learn collects site events into an indicator table, triggers classification after 7 events critical mass, and… | ||
| Keynote: Working Efficiently with Big Data in Text Formats | David Mertz | 44m |
Continuum Analytics open sources the CSV/JSON library TextAdapter from its commercial IOPro suite, adds row-position indexing stored locally, and demonstrates… | ||
| Why is Windows so broken? and what to do about it | Steve Dower | 42m |
Python on Windows is concretely better: over 4 million Windows downloads a month, official Visual C++ compilers matched to Python versions are available, and… | ||
| Similarity Search in Patents via modern Artificial Intelligence | Sumeet Sandhu, Tami Lee | 41m |
Built a production Patent Management Tool that uses Python and Deep Learning to generate data vocabularies for similarity search and shows preliminary… | ||
| Derivation & Presentation: How to Effectively Tell A Data Science Story | Gregory Kamradt | 38m |
Claim: data science succeeds only when it prescribes actionable change, not by running models for their own sake. Time in practice splits into prep/quality… | ||
| Mental Models to Use and Avoid as a Data Scientist | Jonathan Whitmore | 35m |
Uses Jupyter Notebooks with runnable Python to show concrete examples and shares slides and code via a bit.ly link. Covers specific pitfalls (p‑hacking, garden… | ||
| improv BOF How to Scrape Data from any Website | 24m | |
Concrete howtos and tradeoffs: a live curl + grep demo shows pulling meetup member lists, then HTML parsing with BeautifulSoup for quick one-offs, moving to… | ||
| nothing matches | ||