• PyData: The History of Python Development, Reactivating Customers & Pathlib!

    Hallo everyone, Our September meetup in 2019 will be hosted by HelloFresh ( https://www.hellofresh.com/ ). Doors open at 7:00 PM and the talks will start at 7:20 PM. Food and Drinks will be available at 7 PM, and are graciously provided by our hosts, HelloFresh. We have 3 speakers scheduled for the evening: 1. "Reactivating Customers" / Tom Kealy, Data Scientist at HelloFresh 2: "The Pathlib Module! (PEP 428)" (lightning talk) / Imanol Uriarte . Break & Mingling! 3. "The History of Python Development" / Mark Neznansky (abstracts below) Please contact us at [masked] if you would like to share something with the community in a 5 minutes talk. We look forward to see all of you you there! ================================ TALK ABSTRACTS: Talk 1: "Reactivating Customers" (20 min) Speaker: Tom Kealy, Data Scientist at HelloFresh Abstract: As part of the customer journey within HelloFresh, a customer may cancel and reactivate a subscription. Reactivating customers have a lower acquisition cost than acquiring new customers, so periodically HelloFresh will offer former customers discounts to reactivate their account. However, there is a large difference between the most valuable and least valuable reactivating customers. There is also the problem of customers who reactivate only for the deal (‘deal seekers’). This talk aims on how we are working to identify the highest value customers, defined in terms of 10 week revenue, per reactivation campaign. This is to help steer campaigns by only contacting probable high value customers, and to decide which discounts different groups of customers receive. This talk will cover the technical aspects of identifying the high value customers and preliminary results for the May 2019 reactivations campaign. Talk 2: "The Pathlib Module! (PEP 428)" (lightning talk, 10 min) Speaker: Imanol Uriarte Abstract: We'll discuss PEP 428, most likely better known as the pathlib module. It's been a standard library since Python 3.2, but still remains not so widely used. Is it resistance to change? In this talk we will see some of its great features and hopefully some of you will start using it from now on. Link to PEP 428: https://www.python.org/dev/peps/pep-0428/ Bio: Imanol Uriarte is a GeoIT Engineer passionate about data, science, and Python. Talk 3: "The History of Python Development" Speaker: Mark Neznansky Abstract: From a one-person side-project, Python has grown in 30 years to become one of the world's most popular programming languages, dominant among researchers and data scientists. In this talk I will present some of the influences on the language, how it and its hundreds of contributors are coordinated, who can contribute to it and how, who takes the important decisions as well as how the current system of governance came to be. ================================ NumFOCUS Code of Conduct THE SHORT VERSION Be kind to others. Do not insult or put down others. Behave professionally. Remember that harassment and sexist, racist, or exclusionary jokes are not appropriate for NumFOCUS. All communication should be appropriate for a professional audience including people of many different backgrounds. Sexual language and imagery is not appropriate. NumFOCUS is dedicated to providing a harassment-free community for everyone, regardless of gender, sexual orientation, gender identity, and expression, disability, physical appearance, body size, race, or religion. We do not tolerate harassment of community members in any form. Thank you for helping make this a welcoming, friendly community for all. If you haven't yet, please read the detailed version here: https://numfocus.org/code-of-conduct ================================ See you soon! -PyData Berlin Team

  • PyData Berlin August Meetup

    Blacklane

    Hallo everyone, Our August meetup in 2019 will be hosted by Blacklane (https://www.blacklane.com/en/about/). Doors open at 6:30 PM and the talks are going to start at 7:15 PM. Food and Drinks are provided by Blacklane. A Blacklane Ride Voucher will be offered to one participant after a selection at the meetup! We have 2 speakers scheduled for the evening: Talk 1: 30 mins "How the US Military Helps Us Tracking Chauffeurs at Blacklane" David Altmann, Senior Business Intelligence Analyst at Blacklane. GPS data is used for many purposes. At Blacklane we use it to track routes of our Chauffeurs to gain insights in their decisions - often in very traffic-jammed cities. However, raw GPS data is not always ready to use directly. In this talk, we discuss how the raw data can be enhanced using common sense, a Kalman Filter and even a little Machine Learning. Talk 2: 30 mins Fraud Detection in Klarna Vaibhav Singh,Klarna: currently heads the Berlin Data Science team in Klarna with past experience in Credit Risk, Content Moderation who loves to explore new ideas and tech in the ML industry. At Klarna we have been working tirelessly to simplify buying online for over 14 years. Our journey has taken us from a small Swedish payments company to a global player that influences the biggest brands out there. We have 2,500 employees, serve 60 million consumers and work with 100,000+ merchants. Any payment platform will have at some point dealt with fraud, it is just an easy source of money for many people across the world. People try to find loopholes in the system and exploit it as long as it lasts. This talk will highlight some aspects how Klarna fights fraud using Machine Learning at its core and what are some of the lessons learnt from the trenches. Lightning talks "Python Development Process", Mark Neznansky Please contact us at [masked] if you would like to share something with the community in a 5 minutes talk. We look forward to see all of you you there! ---------------------------------------------------------------------------------- NumFOCUS Code of Conduct THE SHORT VERSION Be kind to others. Do not insult or put down others. Behave professionally. Remember that harassment and sexist, racist, or exclusionary jokes are not appropriate for NumFOCUS. All communication should be appropriate for a professional audience including people of many different backgrounds. Sexual language and imagery is not appropriate. NumFOCUS is dedicated to providing a harassment-free community for everyone, regardless of gender, sexual orientation, gender identity, and expression, disability, physical appearance, body size, race, or religion. We do not tolerate harassment of community members in any form. Thank you for helping make this a welcoming, friendly community for all. https://numfocus.org/code-of-conduct ---------------------------------------------------------------------------------- See you soon! -PyData Berlin Team

    6
  • PyData Berlin July Meetup

    Planet Labs Germany GmbH

    Hallo everyone, Our July meetup in 2019 will be hosted by Planet. Doors open at 6:30 PM and the talks are going to start at 7:15 PM. Food and Drinks are provided by Planet. We have 2 speakers scheduled for the evening Talk 1: 35 mins Python in the Mangroves: tracking ecosystem health from space by Rodrigo Almeida, Planet Mangroves forests are critical to coastal ecosystems. Making use of Planet Basemaps we are able to track the health and size of the mangrove area consistently from space. Packages such as Rasterio and GeoPandas are used within Python to create a procedure to generate plots of the evolution of mangrove area and health over time. Similar projects have been developed into production applications for Coral Reef monitoring such as the Allen Coral Atlas. Data is a key component of the SDGs and Earth observation can provide valuable information to solve the world’s toughest challenges. With a background in Agronomic Engineering and a masters degree in Geo-Information Science from Wageningen University, Rodrigo has been deeply involved in developing costumer relations and use cases in Europe with Planets imagery. He moved to Berlin for an internship at Planet where he was at the forefront of our new analytics products and processing capabilities, working closely with the Engineering team to meet costumers needs Talk 2: 35 mins Moving to streaming data processing, an alternative approach by Maximos Sapranidis, Klarna The faster you process data and the more data you can hold, the better it is for optimising your business needs and given the pace of the modern world the typical batch processing paradigm has reached its limitations. This led to a new streaming paradigm, that moves away from fixed time processing into on-demand processing thus allowing for new opportunities and faster feedback cycles. Most of the times those platforms come with their own caveats when it comes to operating and maintaining them and most importantly scaling them. The aim of this presentation is to offer an alternative solution when building a stream processing platform, that reduces the operational costs and allows for an extremely simple yet powerful computing model without the headache of scaling the system. The solution is based around serverless technologies with minimal provisioning and orchestration overhead and it the solution currently utilized inside Klarna for its risk capabilities. Maximos Sapranidis is a passionate software engineer and architect currently working at Klarna in the decision services department. Previous to that Maximos was responsible for a lot of different FinTech products like the peer to peer application of Klarna that originated from Cookies labs and Number26 Lightning talks The wonderful world of PEPs - Matti Lyra Please contact us at [masked] if you would like to share something with the community in a 5 minutes talk. We look forward to see all of you you there! ---------------------------------------------------------------------------------- NumFOCUS Code of Conduct THE SHORT VERSION Be kind to others. Do not insult or put down others. Behave professionally. Remember that harassment and sexist, racist, or exclusionary jokes are not appropriate for NumFOCUS. All communication should be appropriate for a professional audience including people of many different backgrounds. Sexual language and imagery is not appropriate. NumFOCUS is dedicated to providing a harassment-free community for everyone, regardless of gender, sexual orientation, gender identity, and expression, disability, physical appearance, body size, race, or religion. We do not tolerate harassment of community members in any form. Thank you for helping make this a welcoming, friendly community for all. https://numfocus.org/code-of-conduct ---------------------------------------------------------------------------------- See you soon! -PyData Berlin Team

    2
  • PyData June @ babbel

    Babbel | Lesson Nine GmbH

    PyData Berlin June meetup is hosted by babbel. Doors open at 18:45 and the talks are going to start shortly after seven. Food and Drinks are provided generously provided by babbel. We have 2 speakers scheduled for the evening: Daniel Davis and Amit Kushwaha as well as three lightning talks. ---- Using numba along cooperative inheritance to efficiently execute composable simulations by Daniel Davis The talk shows how simcompyl, a discret-time simulation library, applies numba's jit-compilation for blazing fast executions while keeping the complete flexibility of python, including cooperative inheritance and debugging capabilities. Daniel is a Data Scientist working at Gameduell, helping to improve their gaming community site and mobile games experience. He has a strong background in programming and not only enjoys data analysis and model building in python, but also implementing some bigger projects. You may also know him from one of Berlins hackerspaces or from a c3 event, but currently he is quite busy setting up him farm in Havelland. --- Adversarial Machine Learning: Cybersecurity Perspective by Amit Kushwaha Security and Privacy issues need no introduction. But how exactly is this affecting the field of Machine Learning? This is what this talk will cover. We first expose the attack surface of systems deploying machine learning. We then describe how an attacker may force models to make wrong predictions with very little information about the victim. One such attack can be biometric recognition where fake biometric traits may be exploited to impersonate a legitimate user. We demonstrate that these attacks are practical against existing machine learning as a service platform. Towards the end, we will discuss current research to defend models from such attacks. Amit is a Python Backend Engineer in the Pricing and Forecasting Team of Zalando SE. He is currently working on Optimal Discount Recommendation on Large Scale. He worked earlier as a Machine Learning Engineer in Zomato. His major areas of interests are Deep Learning, Recommendation Systems, Natural Language Processing and Data Engineering. He is also an Open Source Enthusiast. He is proficient in Python and has worked with the Tensorflow, Keras, Pyspark, Airflow, Luigi and Pandas libraries to name a few. He dreams to pursue Artificial Intelligence as an independent researcher in future.

    4
  • PyData Berlin May Meetup

    Applift

    Hallo everyone, Our May meetup in 2019 will be hosted by Applift GmbH. Doors open at 6:40 PM and the talks are going to start at 7:15 PM. Food and Drinks are provided. we have 2 speakers scheduled for the evening Talk 1: 35 mins Title: ELT with BigQuery and Python. Pros, Cons and Pitfalls. Speakers: Sowmia Naraynan (Team Lead Business Intelligence) and Cian Cullinan (Business Intelligence Developer) Talk 2: 35 mins Title: The Anaconda Journey and the future of Open Data Science Speaker: Peter Wang, Anaconda Inc Lightning talks 1. Dinu Gherman: “Spheromones” 2. Tania Vasilikioti: `SettingwithCopyWarning` in Pandas Please contact us at [masked] if you would like to share something with the community in a 5 minutes talk. We look forward to see all of you you there! ---------------------------------------------------------------------------------- The CFP is open for the 2019 PyData berlin and PyCon DE awesome conference in October check out more details https://de.pycon.org/ Also check out Pyladies Berlin have a meetup event for Call for Proposals Mentoring https://www.meetup.com/PyLadies-Berlin/events/260199721/ It would be a pleasure to have you giving a talk at one of our meetups, please fill out our handy form with your idea: https://berlin.pydata.org/talk-or-event-idea-submission/ ---------------------------------------------------------------------------------- NumFOCUS Code of Conduct THE SHORT VERSION Be kind to others. Do not insult or put down others. Behave professionally. Remember that harassment and sexist, racist, or exclusionary jokes are not appropriate for NumFOCUS. All communication should be appropriate for a professional audience including people of many different backgrounds. Sexual language and imagery is not appropriate. NumFOCUS is dedicated to providing a harassment-free community for everyone, regardless of gender, sexual orientation, gender identity, and expression, disability, physical appearance, body size, race, or religion. We do not tolerate harassment of community members in any form. Thank you for helping make this a welcoming, friendly community for all. https://numfocus.org/code-of-conduct ---------------------------------------------------------------------------------- See you soon! -PyData Berlin Team

    3
  • PyData Berlin April Meetup

    Wooga

    Hallo everyone, Our April meetup in 2019 will be kindly hosted by Wooga. Doors open at 6:45 PM and the talks are going to start at 7:15 PM. Food and Drinks are provided. So far we have one speaker scheduled and we are looking for a second speaker. Please reach out to us (here, or via [masked]) or fill the form on the website (https://berlin.pydata.org/) If you would like to give a talk. Talk 1: 35 mins Title: Bokeh basics and some advanced features Speaker: Dmitri Stepanov Abstract: It is hard to overestimate the importance of data visualisation. Most of the introductory courses/tutorials on data analysis or example notebooks employ libraries such as matplotlib . But how about something more interactive? What if you want to zoom in or get some additional information on a data point by hovering over it on a plot? I will introduce Bokeh library which is packed with tools for such interactive visualisations in browsers. I will briefly cover the basics and focus on slightly more advanced features such as UI controls like buttons and sliders and embedding Bokeh plots into web applications. Speaker Bio: Dmitri comes originally from the pharmaceutical domain. Quickly realising that his interests lay instead in computer/data science, he started a project in machine learning. Since three years he is a PhD student at the Freie Universität Berlin working on a machine learning approach to predict human skin penetration by chemical compounds. Talk 2: 35 mins TBD Lightning talks 1. Chrome extension against woman harassment Speaker: Teresa Ingram, Pyladies Berlin host. Python was her first love but life led her to working with Java she is now plunging back into the Python world. Please contact us at [masked] if you would like to share something with the community in a 5 minutes talk. We look forward to see all of you you there! ---------------------------------------------------------------------------------- Stay tuned: In 2019, PyData Berlin is going to join forces with PyCon.DE to organise an awesome conference in October: https://de.pycon.org/ It would be a pleasure to have you giving a talk at one of our meetups, please fill out our handy form with your idea: https://berlin.pydata.org/talk-or-event-idea-submission/ ---------------------------------------------------------------------------------- NumFOCUS Code of Conduct THE SHORT VERSION Be kind to others. Do not insult or put down others. Behave professionally. Remember that harassment and sexist, racist, or exclusionary jokes are not appropriate for NumFOCUS. All communication should be appropriate for a professional audience including people of many different backgrounds. Sexual language and imagery is not appropriate. NumFOCUS is dedicated to providing a harassment-free community for everyone, regardless of gender, sexual orientation, gender identity, and expression, disability, physical appearance, body size, race, or religion. We do not tolerate harassment of community members in any form. Thank you for helping make this a welcoming, friendly community for all. https://numfocus.org/code-of-conduct ---------------------------------------------------------------------------------- See you soon! -PyData Berlin Team

    9
  • PyData Berlin March Meetup

    Naspers OLX - Berlin

    Hallo everyone, Our March meetup in 2019 will be hosted by OLX. Doors open at 6:45 PM and the talks are going to start at 7:15 PM. Food and Drinks are provided. The Schedule for the evening: Talk 1: 35 mins Title: Image classification at scale Speaker: Alexey Grigorev Abstract: At OLX.pl people upload 3 mln images per day, at OLX.ua 1.5 mln, and in total across all OLX sites it can easily be tens of millions images daily. It is important for us to know what is on these images right after they are uploaded, and we use deep learning classification models for that. Serving deep learning models is challenging at this scale. In this talk we will present our image processing infrastructure. We will cover all steps of model lifecycle: from how the models are trained to how they are deployed to our kubernetes cluster. Speaker Bio: Alexey Grigorev is a data scientist and machine learning engineer with more than 8 years of professional experience. Currently Alexey works as a data scientist at OLX. His areas of expertise are machine learning and text mining. Talk 2: 35 mins Title: How autonomous driving can be tackled with a single neural network using Python and Tensorflow Speaker: Markus Hinsche Abstract: Deep networks can be trained on demonstrations of human driving to learn to follow roads and avoid obstacles. This is possible with a single end-to-end network learning all the parts of the driving at once. I will give a short introduction to autonomous driving stacks and guide you through the implemention of this network which was introduced in the paper "End-to-end Driving via Conditional Imitation Learning". We open-sourced our implementation (https://github.com/merantix/imitation-learning) and wrote a Medium post (https://medium.com/merantix/journey-from-academic-paper-to-industry-usage-cf57fe598f31). Speaker Bio: Markus Hinsche is a Software Engineer working on Machine Learning at Merantix and one of the rare breed of people actually from the Berlin area. He is eager to explore new topics every day. To satisfy this hunger for the unknown, Markus worked at various startups after receiving his Master's degree in IT Systems Engineering at Hasso Plattner Institute in Potsdam. Lightning talks Please contact us at [masked] if you would like to share something with the community in a 5 minutes talk. We look forward to see all of you you there! ---------------------------------------------------------------------------------- Stay tuned: In 2019, PyData Berlin is going to join forces with PyCon.DE to organise an awesome conference in October. It would be a pleasure to have you giving a talk at one of our meetups, please fill out our handy form with your idea: https://berlin.pydata.org/talk-or-event-idea-submission/ ---------------------------------------------------------------------------------- NumFOCUS Code of Conduct THE SHORT VERSION Be kind to others. Do not insult or put down others. Behave professionally. Remember that harassment and sexist, racist, or exclusionary jokes are not appropriate for NumFOCUS. All communication should be appropriate for a professional audience including people of many different backgrounds. Sexual language and imagery is not appropriate. NumFOCUS is dedicated to providing a harassment-free community for everyone, regardless of gender, sexual orientation, gender identity, and expression, disability, physical appearance, body size, race, or religion. We do not tolerate harassment of community members in any form. Thank you for helping make this a welcoming, friendly community for all. https://numfocus.org/code-of-conduct ---------------------------------------------------------------------------------- See you soon! -PyData Berlin Team

  • PyData Berlin February meetup

    WATTx GmbH

    Hallo everyone, Our Feb meetup in 2019 will be hosted by WATTx GmbH Doors open at 6:30 PM and the talks are going to start at 7:15 PM. Food and Drinks are provided. The Schedule for the evening: Talk 1: 35 mins Title: NLP concepts Speaker: Steph Samson Abstract: You've probably heard the words n-grams, tokens, POS tags and other mysterious terms thrown around during some PyData meetups but what do they exactly mean? In this talk, I will be giving a short introduction to linguistics for natural language processing. By the end of this talk, attendees should be familiar with some basic linguistic terms to help them better understand NLP-related presentations and tools. Speaker Bio: Steph is a software engineer at MOIA, a VW-company, working with the autonomous mobility team. She is trained as a computational linguist and previously worked on natural language processing. You may also know her from the Recurse Center (formerly known as Hacker School). Talk 2: 35 mins Title: The ImageNet Moment in NLP: From wordvectors to language models Speaker: Tobias Sterbak Abstract: The year 2018 was considered the year of pre-training in natural language processing (NLP), where many important benchmarks where dominated by large pre-trained and fine-tuned neural networks. Some already claimed the "ImageNet moment in NLP" has arrived. While in the past, only wordvectors where trained on large corpora of text, the recent development trained full language models and fine-tuned them on a specific task. I will give you an overview of the models and methods to fine-tune them. We will cover ELMo, ULMFit and Bert. After the talk you will know how you can apply the methods in keras and pytorch. Speaker Bio: Tobias Sterbak is a freelance machine learning consultant, providing state-of-the-art natural language processing and machine learning for companies in multiple industries. He is also writing about machine learning and natural language processing on www.depends-on-the-definition.com. Lightning talks Rüdiger Busche: Sacred (https://github.com/IDSIA/sacred) Damian Heimel: AI from research to production, lessens learned in http://www.deevio.ai/ We look forward to see all of you you there! ---------------------------------------------------------------------------------- Stay tuned: In 2019, PyData Berlin is going to join forces with PyCon.DE to organise an awesome conference in October. It would be a pleasure to have you giving a talk at one of our meetups, please fill out our handy form with your idea: https://berlin.pydata.org/talk-or-event-idea-submission/ ---------------------------------------------------------------------------------- NumFOCUS Code of Conduct THE SHORT VERSION Be kind to others. Do not insult or put down others. Behave professionally. Remember that harassment and sexist, racist, or exclusionary jokes are not appropriate for NumFOCUS. All communication should be appropriate for a professional audience including people of many different backgrounds. Sexual language and imagery is not appropriate. NumFOCUS is dedicated to providing a harassment-free community for everyone, regardless of gender, sexual orientation, gender identity, and expression, disability, physical appearance, body size, race, or religion. We do not tolerate harassment of community members in any form. Thank you for helping make this a welcoming, friendly community for all. https://numfocus.org/code-of-conduct ---------------------------------------------------------------------------------- See you soon! -PyData Berlin Team

    7
  • PyData Berlin - Meet & Drink!

    März Restaurant

    Hello PyData Berliners! Again!; in addition to our monthly meetups, we are hosting an after work meet & drink. Why not have some drinks with us after work in these cold dark days? Join us for a relaxed bar evening @ März Restaurant (they have food and drinks, and it's more of a cafe than a restaurant!) We are excited to see you there! Notes: 1. NumFocus's Code of Conduct applies (https://numfocus.org/code-of-conduct) 2. Unlike our usual meetups, food and drinks are not sponsored! 3. Just in case meetup shows the wrong location: https://goo.gl/maps/WL8axvcie6A2

  • PyData Berlin - First meetup in 2019

    Delivery Hero SE

    Hallo everyone, Our first meetup in 2019 will be hosted by DeliveryHero. Doors open at 6:50 PM and the talks are going to start at 7:15 PM. Food and Drinks are provided The Schedule for the evening: Talk 1: 35 mins Title: Can you explain your model? Boosting Domain Knowledge and Decision Making with Model Interpretation Speaker: Iván Lopez Abstract: Modern Machine Learning models behave like black-boxes in which we input values and retrieve some predictions. Models can be evaluated over unseen data using the desired metrics, but this is not enough to gain the trust of business stakeholders, who would like to understand why the model is predicting what it is predicting. Model interpretation sheds light on ML models. As data scientists, it is useful to know what is inside the black box to better understand the problem, identify biases, improve generalization, etc. Equally important, supporting business stakeholders with automatic explanations for every prediction fosters their trust and gains their interest on improving solutions, while also closing the gap between predictions and decision making. In this talk, we share our work at Delivery Hero on extending traditional ML outputs with automatic explanations for decision making. Speaker Bio: Iván is a Machine Learning Engineer. He holds a Master degree in Datascience by the University of Nice and TUB. Since two years, he has been working at Delivery Hero implementing ML models to predict customers' behavior Lightning talk by Eugene: Lessons from the Advent of Code Lightning talk by Marcus Jones: Connecting the dots with JupyterHub and the Ethereum Blockchain There are still open slots for lightning talks, if you're interested in giving one, send us an email at [masked]. We look forward to see all of you you there! ---------------------------------------------------------------------------------- Stay tuned: In 2019, PyData Berlin is going to join forces with PyCon.DE to organise an awesome conference in October. It would be a pleasure to have you giving a talk at one of our meetups, please fill out our handy form with your idea: https://berlin.pydata.org/talk-or-event-idea-submission/ See you soon! -PyData Berlin Team

    13