PyData DC 2016
| title | author | length |
|---|---|---|
| ElasticSearch and Redis How and When to Use Them | Tim Marcinowski | 36m |
Shows concrete recipes: scraping websites and the GitHub API, indexing ~30 users into Elasticsearch, finding each user’s top language and observing ~10 unique… | ||
| Becoming a Data Scientist Advice From My Podcast Guests | Renee Teate | 44m |
Seventeen podcast interviews and a Twitter network of about 9,500 followers produce concrete guidance: start by making a simple project report, learn… | ||
| Design Principles | James Powell | 42m |
Argues that complexity needs concrete justification and offers a usable framework called a complexity budget that trades user sophistication and user care to… | ||
| Bot or Not? The Illusion of Intelligence | Bobby Filar | 40m |
A practical report of what actually works: Python libraries SpaCy, NLTK, sklearn, Keras get wired into a retrieval QA pipeline that ingests a file by SHA256… | ||
| Exposing Algorithms | Jennifer Stark | 35m |
Shows concrete investigations: a Google autocomplete probe that maps words blocked by Google and Bing, and an Uber study that samples DC on a dense grid every… | ||
| Keynote: The Culture of Data Transformation | SriSatish Ambati | 29m |
Claims that matter: concrete engineering and results. Describes that H2O rebuilt its data platform for 1k–10k+ features and redesigned storage for… | ||
| Dynamics in Graph Analysis Adding Time as a Structure for Visual and Statistical | Benjamin Bengfort | 42m |
Claims that embedding time structurally reduces traversal cost and clarifies dynamics are demonstrated with concrete tools and counts: NetworkX and graph_tool… | ||
| Machine Learning Techniques for Class Imbalances & Adversaries | Brendan Herger | 30m |
A practical survey that reports what is actually tried and what works: the Data Innovation Lab tests SMOTE sampling and finds synthetic minority interpolation… | ||
| Visual diagnostics for more informed machine learning | Rebecca Bilbro | 39m |
Argument: adding visual diagnostics at each pipeline stage makes machine learning more informed, not just prettier. She demonstrates concrete tools… | ||
| Dev Ops meets Data Science Taking models from prototype to production with Docker | Andy Terrel | 46m |
Shows a working path from prototype model to production API using Docker, Kubernetes and Anaconda and reports concrete outcomes: Docker image deploy in 10… | ||
| Data Transformation: A Framework for Exploratory Data Analysis | Tony Ojeda | 32m |
Introduces a repeatable exploratory framework and applies it to the EPA vehicle fuel economy dataset, showing concrete steps: identify entities and… | ||
| Triaging Feedback Form Data | Stephanie Kim | 32m |
Uses Python and R to clean and model Intercom feedback, starting from roughly 500 cleaned messages after deduplication and short-message removal. Builds TF-IDF… | ||
| You got your engineering in my Data Science: Addressing the reproducibility crisis | Jon Bodner | 35m |
Argues that adopting software engineering practices actually reduces the reproducibility crisis and cites concrete failures: fMRI pipelines flag spurious… | ||
| H2O Deep Water with Python early sneek | Fabrizio Milo | 42m |
Python code defines networks that run remotely on an H2O cluster using TensorFlow MXNet or Caffe; the demo loads pre‑built architectures (LeNet AlexNet VGG… | ||
| Open Data Dashboards & Python Web Scraping | Marie Whittaker | 40m |
Dashboard ingests and reshapes open data with Python and scrapers, using packages like openpyxl pandas BeautifulSoup and APIs (Fred), then visualizes in… | ||
| Sustainable scrapers | David Eads | 53m |
NPR builds a scraper that actually produces structured data about gun sale listings on Armslist.com and shows the engineering tradeoffs: Python parsing with… | ||
| GraphGen: Conducting Graph Analytics over Relational Databases | Konstantinos Xirogiannopoulos | 41m |
Claims a bolt-on graph extraction layer over SQL databases that avoids ETL, exposes a Domain-specific datalog-like DSL, and returns an in-memory graph object… | ||
| Logistic Regression Behind the Scenes | Christopher White | 31m |
Compares logistic implementations in Python and SAS with concrete examples: scikit‑learn uses liblinear and always regularizes, statsmodels uses Newton/IRLS… | ||
| Improving PySpark Performance Spark performance beyond the JVM | Holden Karau | 41m |
Explains concrete ways to speed PySpark by cutting Python–JVM overhead and shrinking shuffles: uses the py4j bridge, pickling avoidance, RDD cache/persist for… | ||
| Making your code faster: Cython and parallel processing in the Jupyter Notebook | Gustavo Patino | 42m |
Benchmarks run inside a Jupyter Notebook compare plain Python, NumPy, Cython, SciPy integrators, Numba and IPyParallel on an Euler approximation of y=x² over a… | ||
| Dask for ad hoc distributed computing | Matthew Rocklin | 44m |
Describes a pure-Python dynamic task scheduler that executes arbitrary task graphs, coordinates workers and a central scheduler, and measures runtimes and byte… | ||
| Building a (semi) Autonomous Drone with Python | Greg Lamp | 31m |
A working prototype uses a Parrot AR Drone and a nodejs server with a Python client to stream camera frames, run OpenCV and a Python image wrapper, and send… | ||
| A Practical Guide to Dimensionality Reduction Techniques | Vishal Patel | 40m |
Step‑by‑step procedures for reducing 1,933 features down to a usable set with Python code and concrete outcomes: missingness flags, low‑variance drops… | ||
| Agent based modeling in Python | Jackie Kazil | 38m |
Describes a concrete ABM stack: Mesa in Python with examples (Schelling, Wolf‑Sheep), a Tornado‑backed server, a batch runner for parameter sweeps and a data… | ||
| Beyond Bag of Words A Practitioner’s Guide to Advanced NLP | Ariel M’ndange-Pfupfu, Marc Anderson | 46m |
Builds a movie recommender and a time‑series anomaly detector with runnable Python code. Uses the MovieLens 10M dataset, scikit‑learn pipelines and… | ||
| Keynote: Building a Data Driven Dialogue From Filling Potholes to Disrupting the Cycle | Kelly Jin | 34m |
Shows concrete projects and results: a quick Snow Stats prototype and a Google Docs mayor dashboard surfaces 311 photo pipelines updating every 10 seconds; a… | ||
| Building Continuous Learning Systems | Anuj Gupta | 50m |
Builds a global convolutional neural net trained on ~8 million tweets (gives 86% cross‑validation) and lightweight per‑brand online models using a passive… | ||
| Building Serverless Machine Learning Models in the Cloud | Alex Casalboni | 41m |
Builds and deploys actual Python ML models to AWS Lambda using scikit-learn, wires them behind API Gateway, and stores serialized models on S3 so Lambda pulls… | ||
| Clustering: A Guide for the Perplexed | Leland McInnes, John Healy | 35m |
Clustering recommendation: use HDBSCAN rather than kmeans because it outputs noise/outliers, finds variable-density clusters, and avoids forcing every point… | ||
| Creating a Contemporary Lending Risk Management System Using Python | Piero Ferrante | 40m |
A complete lending risk stack built on the PyData ecosystem uses rpy2 for R forecasting and Spyre for a lightweight web UI, running millions of models nightly… | ||
| Data Sciencing While Female | Mandi Traud | 31m |
A Meetup grows to 1,227 members in just over a year, backed by concrete events: workshops, a hackathon on traffic data, a job board and sponsorships that… | ||
| Eat Your Vegetables Data Security for Data Scientists | Will Voorhees | 40m |
Claim: basic hygiene plus a few concrete tools block most attacks while full cryptography is hard. Evidence and examples: Lloyds estimates ~$400B annual loss… | ||
| Forecasting critical food violations at restaurants using open data | Nicole Donnelly | 25m |
Prototype predicts critical restaurant violations using Python with scraped DC inspection pages (selenium), Yelp, Google Places, Forecast IO and Cclickfix… | ||
| Fuzzy Search Algorithms How and When to Use Them | Jiaqi Liu | 33m |
IPython notebook walks through concrete implementations: uses Jellyfish library for Soundex and Levenshtein, queries a local PostgreSQL instance with the… | ||
| How I learned to time travel, or, data pipelining and scheduling with Airflow | Laura Lorenz | 42m |
Adopts Airflow as the scheduler and executor, wires Celery with RabbitMQ or Redis for distributed workers, and builds a plugin called smart_airflow to add S3… | ||
| JupyterLab: Building Blocks for Interactive Computing | Jason Grout | 40m |
JupyterLab claims to be an alpha preview extensible web application that assembles notebooks editors terminals a file browser and a command palette into a… | ||
| Keynote: Become a Data Superhero How Data Can Change the World | Elizabeth Lindsey | 31m |
Byte Back details a concrete intervention: a 24-hour data dive that ingests 3,300 student records and links Excel with Salesforce and Excel exports to run… | ||
| NoSQL doesn't mean No Schema | Steven Lott | 44m |
Claim: NoSQL still has a schema and it belongs in code. Uses Python plus the jsonschema validator and BSON (Pongo) to embed a JSON Schema dictionary in a model… | ||
| Predicting Usage for Capital Bikeshare stations based upon Spatial Characteristics | Darshan Pandit | 21m |
Uses Capital Bikeshare data plus OpenStreetMap and the Smart Location Database to compute station-by-station features: population and job densities, facility… | ||
| Promoting a data driven culture in a world of microservices | Alex DeBrie, Kelly Burdine | 32m |
Hudl consolidates logs and databases into Amazon Redshift and opens warehouse access to all 430 employees, runs 24 TB and 100B+ rows, and spins up data… | ||
| Python useRs | Daniel Chen | 45m |
Practical comparison of workflows and failures: example datasets are Ebola, a survey set, and Gapminder; Python demos use pandas, Jupyter/Anaconda… | ||
| Scaling up to Big Data Devops for Data Science | Marck Vaisman | 35m |
Claims that spinning analytic clusters is doable with scripts but the last 10% of configuration breaks most flows and requires system work; demonstrates using… | ||
| Variational Inference in Python | Austin Rochford | 37m |
Variational inference is presented as an optimization alternative to MCMC with concrete tooling and results: Edward (TensorFlow) and PyMC3 (Theano) run… | ||
| Using Dask for Parallel Computing in Python | Matthew Rocklin | 44m |
Dask provides parallel arrays, bags, and dataframes that build task graphs from many small NumPy or pandas pieces, run on threads, processes, or a distributed… | ||
| The Five Kinds of Python Functions | Steven Lott | 01h 14m |
Clear claims and runnable code: five concrete function types are demonstrated with working examples — diesel-engine RPM case study (a naive model then a fitted… | ||
| Parallel Python Analyzing Large Data Sets | Aron Ahmadia, Matthew Rocklin | 01h 01m |
Shows concrete patterns and results: using concurrent.futures process/thread pools to turn a JSON-to-HDF5 conversion from slow sequential loops into a parallel… | ||
| Pandas from the Inside | Stephen Simmons | 01h 21m |
Slides and code dig into pandas internals using numpy arrays and real AFL game data and build a season ladder that matches the AFL website. Concrete claims… | ||
| Modern NLP in Python | Patrick Harrison | 01h 31m |
Code loads spaCy, gensim and pyLDAvis and runs on the Yelp snapshot of 1 million restaurant reviews to: clean and phrase‑model text, train a 50‑topic LDA… | ||
| Machine Learning with Text in scikit learn | Kevin Markham | 01h 24m |
Builds and evaluates real SMS spam classifiers from a GitHub tutorial using scikit-learn and pandas, turning text into a sparse document‑term matrix with count… | ||
| Learn how to Make Life Easier with Anaconda | Dhavide Aruliah | 01h 22m |
Shows exactly how to reproduce working environments: use conda create to make a Python 2 or R env, run Python 2, Python 3 and R scripts, and fix Unicode and… | ||
| Keynote: How Open Data Science Opens the World of Innovation | Robert Cohn, Peter Wang | 41m |
Claims that up to 20% of servers will run machine learning, that FPGA in-package acceleration arrives for deep learning, and that 3D cross point Optane cuts… | ||
| Julia Tutorial | Chase Coleman | 01h 07m |
Julia claims JIT speed (LLVM) and multiple dispatch and demonstrates concrete tools: Pkg.add/clone, calling C and Python (pycall), Plots with backends like… | ||
| Interactive multi scale time series exploration with matplotlib | Thomas Caswell | 01h 01m |
Shows how to use matplotlib event handling and pick callbacks to build a working multi-scale drilldown viewer that starts from monthly aggregates and drills to… | ||
| How to Build Your Own Self Driving Toy Car | Ryan Zotti | 01h 25m |
Six months of work produce a drivable 12" yellow Ferrari toy car wired with a Raspberry Pi camera and controlled by models built in OpenCV and TensorFlow… | ||
| Getting started with H2O on Python | Ashrith Barthur | 01h 10m |
H2O with Python and optional Hadoop moves log analysis from laptop scripts to a distributed cluster: tutorial shows data pipeline, feature engineering… | ||
| From rocks to a hammer when and how to change your company's analytical tools | Sebastien Genty | 29m |
Company replaces SPSS/SAS (costly at $10k+ per year per license) with Python, IPython/Jupyter and a custom Plato library that automates repeated analytics… | ||
| Educational framework for Black Box optimization methods design | Nadia Udler | 01h 15m |
Shows concrete how-to: uses Scipy.optimize and Scikit-learn notebooks, builds small tutorials on alopi for automatic differentiation, and applies them to… | ||
| Doing frequentist statistics with Scipy | Gustavo Patino | 01h 22m |
Covers concrete commands and gotchas: use SciPy.stats for Kolmogorov Smirnov Shapiro Anderson tests (Anderson returns critical values not a p), t tests and… | ||
| Creating Python Data Pipelines in the Cloud | Femi Anthony | 44m |
Compares Luigi, Airflow and AWS Data Pipeline by building a concrete pipeline that fetches Strava‑style logs, writes raw files to S3, runs ETL into… | ||
| Building Your First Data Pipelines | Hunter Owens | 32m |
Uses Luigi tasks and Targets to turn ETL steps into atomic file or database operations, with a local scheduler for quick runs and a web UI for dependency… | ||
| Beyond Sentiment Emotion Mining with Python and machine learning | Max Tsvetovat | 01h 23m |
Builds a Python pipeline that replaces crude positive/negative scores with a nuanced vector of emotions using Inside Out–style labels, trained on the IMDb… | ||
| Bayesian Network Modeling using R and Python | Pragyansmita Nayak | 41m |
Bayesian networks claim concrete wins: they help locate Airron flight af447 debris, support election models like Nate Sila and R Lindo, and scale to ~1,000,000… | ||
| Closing Session | 21m | |
A lightweight slide tool called sent.py gets created as a quick replacement for Scent with 250 lines of Python and minimal X bindings; an image feature added… | ||
| Keynote: Extending from Open to Usable: A Commerce Data Conundrum | Star Ying | 25m |
Department of Commerce holds roughly a third of federal open datasets and those datasets carry specific pitfalls: five major errata issued for ACS in 2016… | ||
| nothing matches | ||