人工智能
SunJackson
这个作者很懒,什么都没留下…
展开
-
Enhanced text classification and word vectors using Amazon SageMaker BlazingText
Today, we are launching several new features for the Amazon SageMaker BlazingText algorithm. Many downstream natural language processing (NLP) tasks like sentiment analysis, named entity recognit...转载 2018-07-30 10:12:41 · 358 阅读 · 0 评论 -
The Real Problems with Neural Machine Translation
TLDR: No! Your Machine Translation Model is not “prophesying”, but let’s look at the six major issues with neural machine translation (NMT). So I saw a Twitter thread today with the editor-in-ch...转载 2018-07-31 09:18:15 · 208 阅读 · 0 评论 -
Classify your own images using Amazon SageMaker
Image classification and object detection in images are hot topics these days, thanks to a combination of improvements in algorithms, datasets, frameworks, and hardware. These improvements democratize...转载 2018-07-31 09:18:34 · 289 阅读 · 0 评论 -
Partitioning the Variation in Data
Roger Peng ** 2018/07/23One of the fundamental questions that we can ask in any data analysis is, “Why do things vary?” Although I think this is fundamental, I’ve found that it’s not explicitly as...转载 2018-07-31 09:19:20 · 249 阅读 · 0 评论 -
Visualizing an Assassin Puzzle
Over at Math3ma, Tai-Danae Bradley shared the following puzzle, which she also featured in a fantastic (spoiler-free) YouTube video. If you’re seeing this for the first time, watch the video first....转载 2018-07-31 10:54:58 · 239 阅读 · 0 评论 -
Scalable multi-node deep learning training using GPUs in the AWS Cloud
A key barrier to the wider adoption of deep neural networks on industrial-size datasets is the time and resources required to train them. AlexNet, which won the 2012 ImageNet Large Scale Visual Recogn...转载 2018-07-31 10:55:21 · 364 阅读 · 0 评论 -
人工智能将如何在2019年改变零售业(How AI Will Change Brick-and-Mortar Retail in 2019)
Data science, machine learning, and AI have clear applications for e-commerce, and given their relative ease of implementation, most online retailers are already deeply invested in strategies like rec...转载 2018-12-27 09:24:08 · 768 阅读 · 0 评论 -
LSTM的神奇之处
前言LSTM神经网络代表长期短期记忆,是一种特殊类型的递归神经网络,最近在机器学习界引起了很多关注。简而言之,LSTM网络内部具有一些上下文状态单元,它们充当长期或短期存储器单元。LSTM网络的输出由这些单元的状态调制而成。当我们的神经网络需要依赖于输入的历史背景而不是仅仅依赖于最后的输入进行预测时,这是一个非常重要的属性。举个简单的例子,设想我们想要预测一个序列的下一个数字:6 ->...原创 2018-12-27 11:13:00 · 488 阅读 · 0 评论 -
机器学习和深度学习中如何处理数据不平衡问题
如何处理数据不平衡问题前言在您正在处理数据集时您可以创建分类模型并立即获得90%的准确度。你觉得“非常不错”。但是当你深入一点时,发现90%的数据属于一个类。这是一个不平衡数据集的例子,它可能导致令人沮丧的结果。当你发现你的数据有不平衡的类并且你认为你得到的所有好的结果都变成了错误的时候,你会感到非常沮丧。当你发现大部分书籍,文章和博客文章似乎并没有为您提供有关处理数据不平衡的良好建议时...翻译 2019-01-25 11:39:24 · 3843 阅读 · 0 评论 -
特征工程-数据处理
特征工程连续型变量连续变量无量纲化连续变量数据变换连续变量离散化类别变量时间型、日期型变量缺失值处理特征组合连续型变量处理什么是连续型变量?在一定区间内可以任意取值的变量叫连续变量,其数值是连续不断的,相邻两个数值可作无限分割,即可取无限个数值.例如,生产零件的规格尺寸,人体测量的身高,体重,胸围等为连续变量,其数值只能用测量或计量的方法取得.连续变量无量纲化统...原创 2019-03-17 11:35:34 · 1278 阅读 · 0 评论 -
DeepMind提出空间语言集成模型SLIM
选自arXiv,作者:Tiago Ramalho , Tomáš Kociský等,机器之心编译,参与:陈韵竹、路。 前不久,DeepMind 提出生成查询网络 GQN,具备从 2D 画面到 3D 空间的转换能力。近日,DeepMind 基于 GQN 提出一种新模型,可以捕捉空间关系的语义(如 behind、left of 等),其中包含一个基于从场景文本描述来生成场景图像的新型多模态目标...转载 2018-07-27 14:54:57 · 242 阅读 · 0 评论 -
Whats new on arXiv
GPU-based Commonsense Paradigms Reasoning for Real-Time Query Answering and Multimodal Analysis We utilize commonsense knowledge bases to address the problem of real- time multimodal analysis. In par...转载 2018-07-27 14:42:52 · 804 阅读 · 0 评论 -
New Research on Multi-Task Learning
We are excited to share the latest report and prototype from our machine intelligence R team: Multi-Task Learning.Wax on.. face off! When humans learn new tasks, we take advantage of knowledge we’...转载 2018-07-30 10:14:26 · 182 阅读 · 0 评论 -
Enhanced text classification and word vectors using Amazon SageMaker BlazingText
Today, we are launching several new features for the Amazon SageMaker BlazingText algorithm. Many downstream natural language processing (NLP) tasks like sentiment analysis, named entity recognit...转载 2018-07-30 10:14:56 · 374 阅读 · 0 评论 -
Object Detection algorithm now available in Amazon SageMaker
Amazon SageMaker is a fully-managed and highly scalable machine learning (ML) platform that makes it easy build, train, and deploy machine learning models. This is a giant step towards the democratiza...转载 2018-07-30 10:15:27 · 297 阅读 · 0 评论 -
RAIN Project: evolution of the game development dream
July 12, 2018Four years ago, I started Data School to share my data science knowledge and help aspiring data scientists achieve their dreams. Today, I’m excited to share with you a new opportunity t...转载 2018-07-30 10:15:48 · 747 阅读 · 0 评论 -
Why AI Isn’t A Black Box (And Its Business Value)
Artificial intelligence (AI) is a powerful and growing field, but some are still hesitant to employ it, seeing it as a black box. However, if you take a second to step outside the box (see what we did...转载 2018-07-30 10:16:52 · 188 阅读 · 0 评论 -
DeepLearning-Github排行
Pos1 Name Description Language Stars Today Total Stars 1 pytorch Tensors and Dynamic neural networks in Python with strong GPU acceleration C++ 18 17245 2 keras...原创 2018-07-27 14:37:39 · 383 阅读 · 0 评论 -
ACL 2018 Highlights: Understanding Representations and Evaluation in More Challenging Settings
You are using an outdated browser. Please upgrade your browser to improve your experience.I attended the 56th Annual Meeting of the Association for Computational Linguistics (ACL 2018) According to...转载 2018-07-27 14:39:58 · 582 阅读 · 0 评论 -
Parsimonious principle vs integration over all uncertainties
tl;dr If you have bad models, bad priors or bad inference choose the simplest possible model. If you have good models, good priors, good inference, use the most elaborate model for predictions. To mak...转载 2018-07-27 14:40:33 · 150 阅读 · 0 评论 -
Distilled News
New Survey Reveals Businesses Are Bullish on Data Lakes The data lake has long served as a powerful tool for data scientists and data engineers. However, today´s business environment often requires t...转载 2018-07-27 14:41:18 · 177 阅读 · 0 评论 -
智能无线网络的深度学习:一项综合调查
摘要作为一种有前途的机器学习工具,用于处理复杂原始数据的精确模式识别,深度学习(DL)正成为向大规模拓扑和复杂无线电传输的无线网络添加智能的有效方法。DL使用许多神经网络层来实现从高维原始数据中快速提取特征。 它可以用于根据大量网络参数(如延迟,丢失率,链路SNR等)的分析来查找网络动态(如热点,干扰分布,拥塞点,流量瓶颈,频谱可用性等)。 因此,DL可以分析具有多个节点和动态链路质量的极其...翻译 2019-05-24 10:01:12 · 6227 阅读 · 0 评论