classification   10112

« earlier    

fast.ai NLP · Practical NLP
fast text classification with deep neural net models.
from fast.ai
nlp  ml  machinelearning  classification  text 
14 hours ago by tswaterman
Opinion | Europe’s Data Protection Law Is a Big, Confusing Mess - The New York Times
There’s another reason for the regulation’s complexity and ambiguity: What are often framed as legal and technical questions are also questions of values. The European Union’s 28 member states have different historical experiences and contemporary attitudes about data collection. Germans, recalling the Nazis’ deadly efficient use of information, are suspicious of government or corporate collection of personal data; people in Nordic countries, on the other hand, link the collection and organization of data to the functioning of strong social welfare systems.
europe  gdpr  classification 
2 days ago by jomc
Introduction to Decision Tree Learning – Heartbeat
From Kaggle to classrooms, one of the first lessons in machine learning involves decision trees. The reason for the focus on decision trees is that they aren’t very mathematics heavy compared to…
machinelearning  python  algorithms  classification  datascience  data_science  DecisionTree  decisiontrees  machine-learning  ml 
2 days ago by tranqy
Introducing state of the art text classification with universal language models · fast.ai NLP
Inductive transfer learning has greatly impacted computer vision, but existing approaches in NLP still require task-specific modifications and training from scratch. We propose Universal Language Model Fine-tuning (ULMFiT), an effective transfer learning method that can be applied to any task in NLP, and introduce techniques that are key for fine-tuning a language model. Our method significantly outperforms the state-of-the-art on six text classification tasks, reducing the error by 18-24% on the majority of datasets. Furthermore, with only 100 labeled examples, it matches the performance of training from scratch on 100x more data. We open-source our pretrained models and code.
ML  NLP  text  classification  fastai 
6 days ago by foodbaby
The Stray Shopping Cart Project
An overview of the Stray Shopping Project, an art project by Julian Montague that constructs an elaborate taxonomy that allows one to identify stray shopping carts based on the situations in which they are found.
art  shopping  cart  funny  taxonomy  classification 
8 days ago by garrettc
Introducing state of the art text classification with universal language models · fast.ai NLP
This post shows how to classify documents automatically with both higher accuracy and less data requirements than previous approaches. It explains in simple terms: natural language processing; text classification; transfer learning; language modeling; and how this approach brings these ideas together.
algorithms  nlp  machinelearning  classification 
9 days ago by peterb
brightmart/text_classification: all kinds of text classificaiton models and more with deep learning
현재까지 나온 모든 종류의 딥러닝 기반 텍스트 분류 기법 총정리
github  nlp  text  classification 
9 days ago by nezz
Introducing state of the art text classification with universal language models · fast.ai NLP
the code and pre-trained models allow anyone to leverage this new approach to better solve problems such as:

- Finding documents relevant to a legal case;
- Identifying spam, bots, and offensive comments;
- Classifying positive and negative reviews of a product;
- Grouping articles by political orientation;
- …and much more.

In computer vision the success of transfer learning and availability of pre-trained Imagenet models has transformed the field. Many people including entrepreneurs, scienti...
nlp  machine-learning  ai  transfer-learning  ml  fast.ai  classification 
10 days ago by hellsten

« earlier    

related tags

2018  accuracy  active  activity  ai  algorithms  america  andrewgroover  anthropology  anttiaarne  api  arcgis  art  arwa.mahdawi  attention  augumentation  autism  automation  bayesian  bias  bigdata  binary  biology  black-box  botany  canada  cart  cartography  cataloging  categorization  class  classify  clustering  cnn  codes  collections  complexity  computervision  consider:representation  continuous  convolutional  coreml  crypto  cryptocurrencies  da  data-fusion  data-imbalance  data-science  data  data_science  datascience  dataviz  davidneale  dcm  debt  decision-trees  decisiontree  decisiontrees  deep-learning  deeplearning  description  detection  dev  development  direction  discontinuous  doc2vec  dynamic  email  emotional_intelligence  estimator  europe  evolution  evolutionary-algorithms  examples  face  facebook  fair  fast.ai  fastai  feature-importance  feedback  fewshot  filtering  fine-grained  fitness  folklore  framework  funny  ga360  gazetteer  gdpr  gender  generative  genetics  genomes  genre  gini-importance  gis  github  goods  google  guide  heat  history  image-analysis  image  incremental  index  industry  inference  infosec  inls201  innovation  internet  javascript  jsic  keras  language  law  learning  libraries  library  linguistic_change  linguistics  literature  location  logistic  longevity  lstm  machine-learning  machine_learning  machinelearning  malware  mapping  market  mbti  media  metadata  metric  ml  mmh  models  motif  multispecies  museology  museum  naics  naivebayes  natural  naturallanguageprocessing  negative-log-likelihood  ner  netsec  neuralnets  news  nll  nlp  north  nudge-targets  object  oclc  oob-error  open  opencalais  opensource  opinion  optimization  people  performance-measure  places  plants  programming  prosopography  python  questionnaire  rachelehrenberg  radical  random-forest  randomforest  rather-interesting  recognition  reference  research  ronaldlanner  rotation  ruslan  saliency  scholars  science  scoring  security  segmentation  sentimentanalysis  services  shopping  social  society  softmax  software  standards  statistics  stiththompson  story  survey  symbology  system  taxes  taxonomy  tech  techdebt  tensorflow  text  text_classification  to-write-about  todo  token  tracker  trademark  transfer-learning  transfer  transferlearning  transformational  translation  trees  trend  tryme  tutorial  typing  uncertainity  validation  valuation  van_tharp  verbs  video-classification  video  vision  visualization  wikipedia  wood  word2vec  writing  yolo 

Copy this bookmark:



description:


tags: