Unconventional Neural Networks
Generative Model Basics
Generating Pythonic code with Neural Network
Generating with MNIST
Classification Generator Training Attempt
Classification Generator Testing Attempt
Drawing a Number by Request with Generative Model
Deep Dream
Deep Dream Frames
Deep Dream Video
Doing Math with Deep Learning (Addition)
#neuralnetwork
Наш телеграм канал - tglink.me/ai_machinelearning_big_data

🎥 Generative Model Basics - Unconventional Neural Networks p.1
👁 2560 раз ⏳ 851 сек.


🎥 Generating Pythonic code with Neural Network - Unconventional Neural Networks p.2
👁 1 раз ⏳ 1188 сек.
Hello and welcome to part 2 of our series of just poking around with neural networks. In the previous tutorial, we played with a generative model, ...

🎥 Generating with MNIST - Unconventional Neural Networks p.3
👁 1 раз ⏳ 616 сек.
Hello and welcome to part 3 of our series of experimenting with neural networks. In this tutorial, we're going to take the same generative model th...

🎥 Classification Generator Training Attempt - Unconventional Neural Networks p.4
👁 1 раз ⏳ 1138 сек.
Hello and welcome to part 4 of our series of having some fun with neural networks, currently with generative networks. Wher we left off, we're buil...

🎥 Classification Generator Testing Attempt - Unconventional Neural Networks p.5
👁 1 раз ⏳ 726 сек.
Hello and welcome to part 5 of our neural network shenanigans series. Lately, we've been working on doing classification with a generative model. I...

🎥 Drawing a Number by Request with Generative Model - Unconventional Neural Networks p.6
👁 1 раз ⏳ 561 сек.
Hello and welcome to part 6 of our neural network antics. In the previous tutorial, we attempted to use a generative model to generate classes of M...

🎥 Deep Dream - Unconventional Neural Networks p.7
👁 1 раз ⏳ 768 сек.
In this part, we're going to get into deep dreaming in TensorFlow. If you are not familiar with deep dream, it's a method we can use to allow a neu...

🎥 Deep Dream Frames- Unconventional Neural Networks p.8
👁 1 раз ⏳ 576 сек.
In this Tutorial, we cover the idea of taking deep dream images, and using them as frames in a video.

Text tutorials and sample code: https://pyth...

🎥 Deep Dream Video- Unconventional Neural Networks p.9
👁 1 раз ⏳ 537 сек.
We've created many deep dream images up to this point, and now we're looking to convert them to video.

Text tutorials and sample code: https://pyt...

🎥 Doing Math with Deep Learning (Addition)- Unconventional Neural Networks p.10
👁 1 раз ⏳ 933 сек.
Can we do math with deep learning? Let's see how far we can go!

Github specific commit for the seq2seq chatbot used here: https://github.com/danie...

​Object Detection with 10 lines of code

🔗 Object Detection with 10 lines of code
Part 2 of this tutorial for detecting your custom objects is available via this link.

🎥 Signate Cigarette Pack Recognition — Николай Сергиевский
👁 1 раз ⏳ 2070 сек.
Николай Сергиевский рассказывает про опыт участия в соревновании Cigarette Pack Recognition на японской платформе Signate. Николай занял второе место. Из видео вы сможете узнать:
- Описание соревнования по распознаванию сигаретных пачек на полках
- Описание решения второго места
- Как выстроить решение для задачи детектирования, классификации и embedding
- Трюки, которые помогают занимать призовые места

Узнать о текущих соревнованиях можно на сайте http://mltrainings.ru/

Узнать о новых тренировках и видео

​Напиши свою песню за 10 минут (модуль textgenrnn Python3)
#Python

Наш телеграм канал - tglink.me/ai_machinelearning_big_data
Сегодня попробуем обучить свою собственную нейронную сеть, чтобы писала текст для песен. Обучающей выборкой будут тексты группы "Руки Вверх". Ничто не мешает чтобы поменять данные на тексты своих любимых групп. Для извлечения данных с веб-сайтов используем Python3 (модуль BeautifulSoup).

Задача будет состоять в том, чтобы выгрузить данные(тексты) c веб-сайтов а потом на их основе обучить нейронную сеть.

На самом деле, можно разбить работу на 2 этапа:
Этап 1: выгрузить и сохранить тексты песни в удобном формате.
Этап 2: обучить свою собственную нейронную сеть.
https://habr.com/ru/post/464973/

🔗 Напиши свою песню за 10 минут (модуль textgenrnn Python3)
Сегодня попробуем обучить свою собственную нейронную сеть, чтобы писала текст для песен. Обучающей выборкой будут тексты группы "Руки Вверх". Ничто не мешает чт...

🎥 3Blue1Brown & Mathematics | Siraj Raval Podcast #3
👁 1 раз ⏳ 4653 сек.
3Blue1Brown aka Grant Sanderson is an outstanding math animation artist, scholar, & educator. In this episode, I ask him about his workflow, the tools he uses, the nature of consciousness, machine learning, & so much more! Grant studied mathematics at Stanford as an undergraduate, then became a multivariable calculus fellow for KhanAcademy through their talent search program. He started making math animation videos on Youtube in 2015, and has since amassed over 2 million subscribers. It was a wonderful con

​SQL Summer Camp: Joins & Unions | Kaggle

🔗 SQL Summer Camp: Joins & Unions | Kaggle
Welcome back, campers! We're kicking the Advanced SQL course off with Joins and Unions. These techniques let you combine data from across different tables in your database. SUBSCRIBE: https://www.youtube.com/c/kaggle?sub_... About Kaggle: Kaggle is the world's largest community of data scientists. Join us to compete, collaborate, learn, and do your data science work. Kaggle's platform is the fastest way to get started on a new data science project. Spin up a Jupyter notebook with a single click. Build wit

​Decoding the Black Box: An Important Introduction to Interpretable Machine Learning Models in Python

🔗 Decoding the Black Box: An Important Introduction to Interpretable Machine Learning Models in Python
Overview

​Exploring Weight Agnostic Neural Networks

http://ai.googleblog.com/2019/08/exploring-weight-agnostic-neural.html

🔗 Exploring Weight Agnostic Neural Networks
Posted by Adam Gaier, Student Researcher and David Ha, Staff Research Scientist, Google Research, Tokyo When training a neural network ...

🎥 Machine Learning with Containers and Amazon SageMaker - AWS Online Tech Talks
👁 1 раз ⏳ 2986 сек.
Data scientists and machine learning engineers use containers to create custom, lightweight environments to train and serve models at scale with deep learning frameworks such as TensorFlow, Apache MXNet, and PyTorch. With containers, developers get consistent environments for development and deployment. In this tech talk, we'll show you how to use AWS Deep Learning Containers to train and serve models at scale with Amazon SageMaker.

Learning Objectives:
- Learn about machine learning using containers and

🎥 Метаобучение в AUTOML: Как строить модели быстрее?
👁 1 раз ⏳ 1919 сек.
Мероприятие: MEETUP день 1
Дата проведения: 23.05.2019

Раскрыта тема доклада:
Мета-обучение в AUTOML.
Как строить модели быстрее?

Спикер: Рыжков Александр
Лаборатория AI Сбербанка

​Money Machines: An Interview With an Anonymous Algorithmic Trader

🔗 Money Machines: An Interview With an Anonymous Algorithmic Trader
An insider explains how algorithms are rewiring finance

🎥 How AI and machine learning will make hoteliers more money while making guests happy!
👁 1 раз ⏳ 1651 сек.
Hoteliers have been talking about collecting data for years, but have struggled to turn it into something practical. Now, the advances in technology and processing power are allowing the use of machine learning to leverage that data in meaningful ways. But, what does that really mean? Kelly McGuire sits down with Glenn and Estella while at HITEC to break all down for us.

Guest: Kelly McGuire, PhD, Revenue Management, Cornell and former Sr VP, Revenue Management at MGM

Sponsors: Cendyn and GCommerce

​Artificial Intelligence Approaches

Authors: Yingjie Hu, Wenwen Li, Dawn Wright, Orhun Aydin, Daniel Wilson, Omar Maher, Mansour Raad

Abstract: Artificial Intelligence (AI) has received tremendous attention from academia, industry, and the general public in recent years. The integration of geography and AI, or GeoAI, provides novel approaches for addressing a variety of problems in the natural environment and our human society.

https://arxiv.org/abs/1908.10345

🔗 Artificial Intelligence Approaches
Artificial Intelligence (AI) has received tremendous attention from academia, industry, and the general public in recent years. The integration of geography and AI, or GeoAI, provides novel approaches for addressing a variety of problems in the natural environment and our human society. This entry briefly reviews the recent development of AI with a focus on machine learning and deep learning approaches. We discuss the integration of AI with geography and particularly geographic information science, and present a number of GeoAI applications and possible future directions.

🎥 Kaggle Gendered Pronoun Resolution — Павел Петроченко, Денис Денисенко
👁 1 раз ⏳ 1426 сек.
Денис Денисенко и Павел Петроченко рассказывают про опыт участия в соревновании Kaggle Gendered Pronoun Resolution, где они заработали серебряную медаль. В команде также участвовали Константин Свиридов, Павел Плесков.

Когда мы читаем текст и встречаем местоимения, то легко понимаем, к какому существительному оно относится. Про эту нетривиальную для компьютера задачу прошёл конкурс, которому и посвящён этот доклад.

Узнать о текущих соревнованиях можно на сайте http://mltrainings.ru/

Узнать о новых трен

​Debugging machine learning models

🔗 Debugging machine learning models
Part 1: Solving high bias and high variance

​Intel выпустил свой первый чип с искусственным интеллектом для дата-центров в Израиле
https://habr.com/ru/news/t/464903/

🔗 Intel выпустил свой первый чип с искусственным интеллектом для дата-центров в Израиле
Центр Intel в Хайфе Intel сообщил новость, которую все давно ждали — компания начала выпускать свой первый чип с искусственным интеллектом, лучший, по соотноше...

​Kaggle Coffee Chat: Joel Grus | Kaggle

🔗 Kaggle Coffee Chat: Joel Grus | Kaggle
In this Coffee Chat Rachael talks with Joel Grus about software engineering best practices, whether they belong in data science, if you should use TensorFlow for fizzbuzz and, of course, why he doesn't like notebooks. You can follow Joel at https://twitter.com/joelgrus SUBSCRIBE: https://www.youtube.com/c/kaggle?sub_confirmation=1&utm_medium=youtube&utm_source=channel&utm_campaign=yt-sub About Kaggle: Kaggle is the world's largest community of data scientists. Join us to compete, collaborate, learn, and

​Matrix processing with nanophotonics

🔗 Matrix processing with nanophotonics
Explanation of how to accelerate deep learning with photonic processors with comparisons to current digital electronics approaches.

​A collection of datasets ready to use with TensorFlow

https://github.com/tensorflow/datasets

🔗 tensorflow/datasets
A collection of datasets ready to use with TensorFlow - tensorflow/datasets

🎥 Running our Reinforcement Learning Agent - Self-driving cars with Carla and Python p.5
👁 1 раз ⏳ 2376 сек.
Now that we've got our environment and agent, we just need to add a bit more logic to tie these together, which is what we'll be doing next to run our reinforcement learning self-driving agent.

Text-based tutorial and sample code: https://pythonprogramming.net/reinforcement-learning-self-driving-autonomous-cars-carla-python/

Channel membership: https://www.youtube.com/channel/UCfzlCWGWYyIQ0aLC5w48gBQ/join
Discord: https://discord.gg/sentdex
Support the content: https://pythonprogramming.net/support-donate

​Deep Reinforcement Learning for Chatbots Using Clustered Actions and Human-Likeness Rewards

Authors: Heriberto Cuayáhuitl, Donghyeon Lee, Seonghan Ryu, Sungja Choi, Inchul Hwang, Jihie Kim
Наш телеграм канал - tglink.me/ai_machinelearning_big_data
Abstract: Training chatbots using the reinforcement learning paradigm is challenging due to high-dimensional states, infinite action spaces and the difficulty in specifying the reward function. We address such problems using clustered actions instead of infinite actions, and a simple but promising reward function based on human-likeness scores derived from human-human dialogue data. We train Deep Reinforcement Learning (DRL) agents using chitchat data in raw text---without any manual annotations.

https://arxiv.org/abs/1908.10331

🔗 Deep Reinforcement Learning for Chatbots Using Clustered Actions and Human-Likeness Rewards
Training chatbots using the reinforcement learning paradigm is challenging due to high-dimensional states, infinite action spaces and the difficulty in specifying the reward function. We address such problems using clustered actions instead of infinite actions, and a simple but promising reward function based on human-likeness scores derived from human-human dialogue data. We train Deep Reinforcement Learning (DRL) agents using chitchat data in raw text---without any manual annotations. Experimental results using different splits of training data report the following. First, that our agents learn reasonable policies in the environments they get familiarised with, but their performance drops substantially when they are exposed to a test set of unseen dialogues. Second, that the choice of sentence embedding size between 100 and 300 dimensions is not significantly different on test data. Third, that our proposed human-likeness rewards are reasonable for training chatbots as long as they use lengthy dialogue hist