coding earth global meetup #5
The 5th coding earth meetup will start with one inspiring talk on Data Science and Machine Learning, followed by a short panel on that topic and an additional gathering to live code and AMA with the speaker. Everyone's invited to contribute code, questions and additional insights. If you're interested in joining our panel, send us a short notice using the contact form at the bottom or use our official email address you find in our imprint.
Here are some of the pointers that have been mentioned during the meetup session for you to follow:
Andrew Ng's online course on machine learning on Coursera
Apache Spark is a data processing library
keras is a leading, well known deep learning framework
tensorboard visualizates neural networks
scikit-learn is one of the most well known machine learning libraries for Python
Tensorflow is the other one, powered by Google but of course free to use
Jupyter Notebooks are the the foundational Python tooling to build interactive playgrounds to play and interact with data (also used in Florians presentation). Google Drive is offering a fully cloud hosted version of them in their Colaboratory Service that lets you start without any setup at all.
xgboost is one of the models that Heike uses at XING and that won many kaggle competitions
kaggle is a great general source for datasets and a stage to have a look how other people solved problems. They're also offering competitions to build models on existing data sets for scientists to compete on (and win prizes)
GPT-3 is an artificial, language understanding intelligence that's smarter than your dog (and might render web designers jobless, if it continues developing at that pace)
Genetic Algorithms and Evolutionary Algorithms are special classes of optimiziation solvers that are good in finding suboptima that analytic optimizers never find.
Generative Adversarial Networks are a special class of combined ML models that interact with each other to somewhat creative results (e.g. create music, paintings or deep fakes)
When you reverse the data flow of certain neural networks they disclose their parameters in fractal looking, hallucinogenic imagery that looks like it originates of a dream.
A classical animal or food example that's hard to solve for most common image recognition models.
Lately, everyone talks about Artificial Intelligence, Machine Learning (ML), Deep Learning… When this hype started, Fabian got a little stressed by all these new terms. What does it all mean for developers? In this talk, Fabian will end the confusion and look at some of the ML fundamentals. We'll be learning the fundamentals by building a simple sentiment classifier. The goal is not to make you a data scientist, but to teach you some of the ML fundamentals to give you an idea of which kind of problems might be solved using ML.
We're welcoming Maike (Porsche), Heike (XING) and Fabian to a virtual ~30 min panel discussion about their experience and learnings over the years. They all work as AI / data science specialists in their fields and we try to keep it as hands on as possible:
- which tools or frameworks would they recommend to get started with?
- what are they currently working on?
- which were there worst pains while becoming AI professionals?
- what are the most exciting developments aside the mainstream and which buzzwords are worth following?
All contributions follow our golden "*1 line of code*" rule, so demos and live code will be abound. You don't have to register to get into our stream but if you do, we send you updates (and nothing else) like the final YouTube URL of the stream before the event.