ailearning

AiLearning：数据分析+机器学习实战+线性代数+PyTorch+NLTK+TF2

39,065

11,403

39,065

View on GitHub

Top Related Projects

ML-For-Beginners

68,986

12 weeks, 26 lessons, 52 quizzes, classic Machine Learning for all

handson-ml

25,189

⛔️ DEPRECATED – See https://github.com/ageron/handson-ml3 instead.

python-machine-learning-book

12,244

The "Python Machine Learning (1st edition)" book code repository and info resource

TensorFlow-Examples

43,354

TensorFlow Tutorial and Examples for Beginners (support TF v1 & v2)

Machine Learning From Scratch. Bare bones NumPy implementations of machine learning models and algorithms with a focus on accessibility. Aims to cover everything from linear regression to deep learning.

ML-YouTube-Courses

15,878

📺 Discover the latest machine learning / AI courses on YouTube.

Quick Overview

The apachecn/ailearning GitHub repository is a comprehensive collection of machine learning and artificial intelligence resources, including tutorials, code examples, and reference materials. It serves as a valuable resource for both beginners and experienced practitioners in the field of AI and machine learning.

Pros

Comprehensive Content: The repository covers a wide range of topics, from fundamental machine learning concepts to advanced techniques and applications.
Multilingual Support: The materials are available in multiple languages, including English, Chinese, and others, making it accessible to a global audience.
Active Community: The project has a vibrant community of contributors, ensuring regular updates and improvements to the content.
Practical Examples: The repository includes numerous code examples and hands-on tutorials, allowing learners to apply the concepts they've learned.

Cons

Uneven Quality: As the content is contributed by a large community, the quality and depth of the materials may vary across different sections.
Lack of Structured Curriculum: The repository is organized as a collection of resources, rather than a structured curriculum, which may make it challenging for beginners to navigate.
Potential Outdated Content: Given the rapid pace of advancements in AI and machine learning, some of the content may become outdated over time.
Language Barriers: While the materials are available in multiple languages, learners who are not proficient in the available languages may face difficulties.

Code Examples

The apachecn/ailearning repository contains a wide range of code examples and tutorials covering various machine learning and AI topics. Here are a few examples:

Linear Regression:

import numpy as np
from sklearn.linear_model import LinearRegression

# Generate sample data
X = np.array([[1, 1], [1, 2], [2, 2], [2, 3]])
y = np.array([5, 8, 9, 11])

# Create and train the linear regression model
model = LinearRegression()
model.fit(X, y)

# Make a prediction
print(model.predict([[3, 5]]))

This code demonstrates the use of the LinearRegression model from the scikit-learn library to perform linear regression on a simple dataset.

K-Means Clustering:

import numpy as np
from sklearn.cluster import KMeans
import matplotlib.pyplot as plt

# Generate sample data
X = np.array([[1, 2], [1, 4], [1, 0], [4, 2], [4, 4], [4, 0]])

# Create and train the K-Means model
model = KMeans(n_clusters=2)
model.fit(X)

# Visualize the clustering results
plt.scatter(X[:, 0], X[:, 1], c=model.labels_, cmap='viridis')
plt.scatter(model.cluster_centers_[:, 0], model.cluster_centers_[:, 1], color='red')
plt.show()

This code demonstrates the use of the KMeans model from the scikit-learn library to perform K-Means clustering on a simple 2D dataset and visualize the results.

Convolutional Neural Network (CNN) for Image Classification:

import tensorflow as tf
from tensorflow.keras.models import Sequential
from tensorflow.keras.layers import Conv2D, MaxPooling2D, Flatten, Dense

# Load and preprocess the dataset
(X_train, y_train), (X_test, y_test) = tf.keras.datasets.mnist.load_data()
X_train = X_train.reshape(-1, 28, 28, 1) / 255.0
X_test = X_test.reshape(-1, 28, 28, 1) / 255.0

# Create the CNN model
model = Sequential([
    Conv2D(32, (3, 3), activation='relu', input_shape=(28, 28, 1)),
    MaxPooling2D((2, 2)),
    Conv2D(64, (3, 3), activation='relu'),
    MaxPooling2D((2, 2)),
    Flatten(),
    Dense(64, activation='relu'),
    Dense

Competitor Comparisons

ML-For-Beginners

68,986

12 weeks, 26 lessons, 52 quizzes, classic Machine Learning for all

Pros of ML-For-Beginners

More structured curriculum with clear learning paths
Extensive documentation and explanations for each concept
Multi-language support for code examples

Cons of ML-For-Beginners

Less focus on advanced topics and cutting-edge techniques
Fewer practical projects and real-world applications

Code Comparison

ML-For-Beginners:

from sklearn.model_selection import train_test_split
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2, random_state=42)

ailearning:

from sklearn.model_selection import train_test_split
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.3, random_state=0)

Both repositories use similar code for splitting datasets, with minor differences in parameters.

Summary

ML-For-Beginners offers a more structured approach to learning machine learning, with clear documentation and multi-language support. However, it may lack depth in advanced topics. ailearning provides a broader range of topics and practical applications but may be less organized for beginners. Both repositories use similar code structures for common machine learning tasks.

handson-ml

25,189

⛔️ DEPRECATED – See https://github.com/ageron/handson-ml3 instead.

Pros of handson-ml

More comprehensive coverage of machine learning topics
Better organized with clear chapter structure
Includes Jupyter notebooks for interactive learning

Cons of handson-ml

Less focus on deep learning and neural networks
Fewer practical examples for real-world applications

Code Comparison

handson-ml:

from sklearn.ensemble import RandomForestClassifier

forest_clf = RandomForestClassifier(n_estimators=100, random_state=42)
forest_clf.fit(X_train, y_train)
y_pred = forest_clf.predict(X_test)

ailearning:

import torch
import torch.nn as nn

class Net(nn.Module):
    def __init__(self):
        super(Net, self).__init__()
        self.conv1 = nn.Conv2d(1, 10, kernel_size=5)
        self.conv2 = nn.Conv2d(10, 20, kernel_size=5)
        self.fc1 = nn.Linear(320, 50)
        self.fc2 = nn.Linear(50, 10)

Summary

handson-ml provides a more structured approach to learning machine learning concepts, with a focus on scikit-learn and traditional ML algorithms. ailearning offers a broader range of topics, including deep learning and neural networks, using frameworks like PyTorch. While handson-ml excels in organization and clarity, ailearning provides more diverse and advanced examples for those interested in cutting-edge AI techniques.

python-machine-learning-book

12,244

The "Python Machine Learning (1st edition)" book code repository and info resource

Pros of python-machine-learning-book

More focused on machine learning concepts and implementations
Provides comprehensive code examples and explanations
Regularly updated with new content and improvements

Cons of python-machine-learning-book

Limited coverage of deep learning and neural networks
Less diverse range of AI topics compared to ailearning
Primarily in English, which may limit accessibility for non-English speakers

Code Comparison

python-machine-learning-book:

from sklearn.model_selection import train_test_split
X_train, X_test, y_train, y_test = train_test_split(
    X, y, test_size=0.3, random_state=1, stratify=y)

ailearning:

import numpy as np
import pandas as pd
from sklearn.model_selection import train_test_split

X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.3, random_state=42)

Both repositories provide code examples for machine learning tasks, but python-machine-learning-book tends to offer more detailed explanations and context for each code snippet. The ailearning repository covers a broader range of AI topics and includes content in multiple languages, making it more accessible to a diverse audience. However, python-machine-learning-book is more focused on machine learning specifically and provides a more structured learning path for this subject.

TensorFlow-Examples

43,354

TensorFlow Tutorial and Examples for Beginners (support TF v1 & v2)

Pros of TensorFlow-Examples

More focused on TensorFlow-specific examples and tutorials
Cleaner, more organized repository structure
Regularly updated with newer TensorFlow versions and features

Cons of TensorFlow-Examples

Limited to TensorFlow framework only
Less comprehensive coverage of general AI/ML concepts
Fewer explanations and theoretical background

Code Comparison

TensorFlow-Examples:

import tensorflow as tf

# Create a constant tensor
hello = tf.constant('Hello, TensorFlow!')

# Start a TensorFlow session
sess = tf.Session()

# Run the op
print(sess.run(hello))

AILearning:

import numpy as np
import pandas as pd
from sklearn.model_selection import train_test_split
from sklearn.linear_model import LinearRegression

# Load and prepare data
data = pd.read_csv('data.csv')
X = data[['feature1', 'feature2']]
y = data['target']

# Split data and train model
X_train, X_test, y_train, y_test = train_test_split(X, y, test_size=0.2)
model = LinearRegression().fit(X_train, y_train)

The code comparison shows that TensorFlow-Examples focuses on TensorFlow-specific code, while AILearning covers a broader range of libraries and techniques in machine learning.

ML-From-Scratch

23,910

Pros of ML-From-Scratch

Focuses on implementing machine learning algorithms from scratch, providing a deeper understanding of the underlying mechanics
Clear and concise Python implementations with minimal dependencies
Includes a wide range of algorithms, from basic to advanced

Cons of ML-From-Scratch

Less comprehensive in terms of overall AI/ML topics compared to AILearning
Lacks extensive documentation and explanations for each algorithm
May not cover the latest cutting-edge techniques in the field

Code Comparison

ML-From-Scratch (Linear Regression implementation):

class LinearRegression(Regression):
    def fit(self, X, y):
        X = np.insert(X, 0, 1, axis=1)
        self.w = np.linalg.inv(X.T.dot(X)).dot(X.T).dot(y)

AILearning (Linear Regression implementation):

def fit_normal(X, y):
    X = np.insert(X, 0, 1, axis=1)
    w = np.linalg.inv(X.T @ X) @ X.T @ y
    return w

Both repositories provide implementations of machine learning algorithms, but ML-From-Scratch focuses more on building algorithms from the ground up, while AILearning offers a broader range of AI and machine learning topics with more extensive documentation and resources.

ML-YouTube-Courses

15,878

📺 Discover the latest machine learning / AI courses on YouTube.

Pros of ML-YouTube-Courses

Curated list of high-quality, free ML courses from YouTube
Organized by topics and skill levels for easy navigation
Regularly updated with new content and community contributions

Cons of ML-YouTube-Courses

Limited to video content only, lacking hands-on exercises or projects
May not cover all AI/ML topics as comprehensively as ailearning
Dependent on external YouTube links, which may become unavailable

Code Comparison

ML-YouTube-Courses doesn't contain code samples, while ailearning includes practical examples. Here's a snippet from ailearning:

# Example from ailearning
import numpy as np

def sigmoid(x):
    return 1 / (1 + np.exp(-x))

def sigmoid_derivative(x):
    return x * (1 - x)

ML-YouTube-Courses focuses on organizing and presenting course information:

## Machine Learning

### Beginner
- [Machine Learning — Andrew Ng, Stanford University](https://www.youtube.com/playlist?list=PLLssT5z_DsK-h9vYZkQkYNWcItqhlRJLN)

Both repositories serve different purposes: ML-YouTube-Courses as a curated list of video resources, and ailearning as a comprehensive AI/ML learning platform with code examples and explanations.

Convert designs to code with AI

Introducing Visual Copilot: A new AI model to turn Figma designs to high quality code using your components.

Try Visual Copilot

README

AI learning

åè®®ï¼CC BY-NC-SA 4.0

ä¸ç§æ°ææ¯ä¸æ¦å¼å§æµè¡ï¼ä½ è¦ä¹åä¸åè·¯æºï¼è¦ä¹æä¸ºéºè·¯ç³ãââStewart Brand

å¨çº¿éè¯»
å¨çº¿éè¯»ï¼v1ï¼
QuantLearning
ApacheCN ä¸æç¿»è¯ç» 713436582
ApacheCN å¦ä¹ èµæº
æ³¨: å¹¿åä½åä½(ç©ç¾ä»·å»)ï¼è¯·èç³» apachecn@163.com

è·¯çº¿å¾

å¥é¨åªç: æ¥éª¤ 1 => 2 => 3ï¼ä½ å¯ä»¥å½å¤§çï¼
ä¸çº§è¡¥å - èµæåº: https://github.com/apachecn/ai-roadmap

è¡¥å

ç®æ³å·é¢: https://www.ixigua.com/pseries/6822642486343631363/
é¢è¯æ±è: https://www.ixigua.com/pseries/6822563009391493636/
æºå¨å¦ä¹ å®æ: https://www.ixigua.com/pseries/6822816341615968772/
NLPæå¦è§é¢: https://www.ixigua.com/pseries/6828241431295951373/
AIå¸¸ç¨å½æ°è¯´æ: https://github.com/apachecn/AiLearning/tree/master/AIå¸¸ç¨å½æ°è¯´æ.md

1.æºå¨å¦ä¹ - åºç¡

æ¯æçæ¬

Version	Supported
3.6.x	:x:
2.7.x	:white_check_mark:

æ³¨æäºé¡¹:

æºå¨å¦ä¹ å®æ: ä»ä»åªæ¯å¦ä¹ ï¼è¯·ä½¿ç¨ python 2.7.x çæ¬ ï¼3.6.x åªæ¯ä¿®æ¹äºé¨åï¼

åºæ¬ä»ç»

èµææ¥æº: Machine Learning in Action(æºå¨å¦ä¹ å®æ-ä¸ªäººç¬è®°)
ç»ä¸æ°æ®å°å: https://github.com/apachecn/data
- ç¾åº¦äºæåå°å: https://github.com/apachecn/data/issues/3
ä¹¦ç±ä¸è½½å°å: https://github.com/apachecn/data/tree/master/book
æºå¨å¦ä¹ ä¸è½½å°å: https://github.com/apachecn/data/tree/master/æºå¨å¦ä¹
æ·±åº¦å¦ä¹ æ°æ®å°å: https://github.com/apachecn/data/tree/master/æ·±åº¦å¦ä¹
æ¨èç³»ç»æ°æ®å°å: https://github.com/apachecn/data/tree/master/æ¨èç³»ç»
è§é¢ç½ç«: ä¼é· ï¼bilibili / Acfun / ç½æäºè¯¾å ï¼å¯ç´æ¥å¨çº¿ææ¾ãï¼æä¸æ¹æç¸åºé¾æ¥ï¼
-- æ¨è çº¢è²ç³å¤´: å°æ¹¾å¤§å¦æè½©ç°æºå¨å¦ä¹ ç¬è®°
-- æ¨è æºå¨å¦ä¹ ç¬è®°: https://feisky.xyz/machine-learning

å¦ä¹ ææ¡£

æ¨¡å	ç« è	ç±»å	è´è´£äºº(GitHub)	QQ
æºå¨å¦ä¹ å®æ	ç¬¬ 1 ç« : æºå¨å¦ä¹ åºç¡	ä»ç»	@æ¯çº¢å¨	1306014226
æºå¨å¦ä¹ å®æ	ç¬¬ 2 ç« : KNN è¿é»ç®æ³	åç±»	@å°¤æ°¸æ±	279393323
æºå¨å¦ä¹ å®æ	ç¬¬ 3 ç« : å³çæ	åç±»	@æ¯æ¶	844300439
æºå¨å¦ä¹ å®æ	ç¬¬ 4 ç« : æ´ç´ è´å¶æ¯	åç±»	@wnma3mz @åæ	1003324213 244970749
æºå¨å¦ä¹ å®æ	ç¬¬ 5 ç« : Logisticåå½	åç±»	@å¾®ååå°	529925688
æºå¨å¦ä¹ å®æ	ç¬¬ 6 ç« : SVM æ¯æåéæº	åç±»	@çå¾·çº¢	934969547
ç½ä¸ç»ååå®¹	ç¬¬ 7 ç« : éææ¹æ³ï¼éæºæ£®æå AdaBoostï¼	åç±»	@çå»	529815144
æºå¨å¦ä¹ å®æ	ç¬¬ 8 ç« : åå½	åå½	@å¾®ååå°	529925688
æºå¨å¦ä¹ å®æ	ç¬¬ 9 ç« : æ åå½	åå½	@å¾®ååå°	529925688
æºå¨å¦ä¹ å®æ	ç¬¬ 10 ç« : K-Means èç±»	èç±»	@å¾ææ¸	827106588
æºå¨å¦ä¹ å®æ	ç¬¬ 11 ç« : å©ç¨ Apriori ç®æ³è¿è¡å³èåæ	é¢ç¹é¡¹é	@åæµ·é£	1049498972
æºå¨å¦ä¹ å®æ	ç¬¬ 12 ç« : FP-growth é«æåç°é¢ç¹é¡¹é	é¢ç¹é¡¹é	@ç¨å¨	842725815
æºå¨å¦ä¹ å®æ	ç¬¬ 13 ç« : å©ç¨ PCA æ¥ç®åæ°æ®	å·¥å·	@å»ç«å¨	835670618
æºå¨å¦ä¹ å®æ	ç¬¬ 14 ç« : å©ç¨ SVD æ¥ç®åæ°æ®	å·¥å·	@å¼ ä¿ç	714974242
æºå¨å¦ä¹ å®æ	ç¬¬ 15 ç« : å¤§æ°æ®ä¸ MapReduce	å·¥å·	@wnma3mz	1003324213
Mlé¡¹ç®å®æ	ç¬¬ 16 ç« : æ¨èç³»ç»ï¼å·²è¿ç§»ï¼	é¡¹ç®	æ¨èç³»ç»ï¼è¿ç§»åå°åï¼
ç¬¬ä¸æçæ»ç»	2017-04-08: ç¬¬ä¸æçæ»ç»	æ»ç»	æ»ç»	529815144

ç½ç«è§é¢

ç¥ä¹é®ç-çç¸å¦-æºå¨å¦ä¹ è¯¥æä¹å¥é¨ï¼

è§é¢æä¹çï¼

çè®ºç§çåºèº«-å»ºè®®å»å¦ä¹ Andrew Ng çè§é¢ï¼Ng çè§é¢ç»å¯¹æ¯æå¨ï¼è¿ä¸ªæ¯åº¸ç½®çï¼
ç¼ç è½åå¼º - å»ºè®®çæä»¬çãæºå¨å¦ä¹ å®æ-æå¦çã
ç¼ç è½åå¼± - å»ºè®®çæä»¬çãæºå¨å¦ä¹ å®æ-è®¨è®ºçãï¼ä¸è¿å¨ççè®ºçæ¶åï¼ç æå¦ç-çè®ºé¨åï¼è®¨è®ºççåºè¯å¤ªå¤ï¼ä¸è¿å¨è®²è§£ä»£ç çæ¶åæ¯ä¸è¡ä¸è¡è®²è§£çï¼æä»¥ï¼æ ¹æ®èªå·±çéæ±ï¼èªç±çç»åã

ãåè´¹ãæ°å¦æå¦è§é¢ - å¯æ±å¦é¢ å¥é¨ç¯

@äºæ¯æ¢ æ¨è: å¯æ±å¦é¢-ç½æå¬å¼è¯¾

æ¦ç	ç»è®¡	çº¿æ§ä»£æ°
å¯æ±å¦é¢(æ¦ç)	å¯æ±å¦é¢(ç»è®¡å¦)	å¯æ±å¦é¢(çº¿æ§ä»£æ°)

æºå¨å¦ä¹ è§é¢ - ApacheCN æå¦ç


AcFun	Bç«

ä¼é·	ç½æäºè¯¾å

ãåè´¹ãæºå¨/æ·±åº¦å¦ä¹ è§é¢ - å´æ©è¾¾

æºå¨å¦ä¹	æ·±åº¦å¦ä¹
å´æ©è¾¾æºå¨å¦ä¹	ç¥ç»ç½ç»åæ·±åº¦å¦ä¹

2.æ·±åº¦å¦ä¹

æ¯æçæ¬

Version	Supported
3.6.x	:white_check_mark:
2.7.x	:x:

å¥é¨åºç¡

Pytorch - æç¨

-- å¾æ´æ°

TensorFlow 2.0 - æç¨

-- å¾æ´æ°

ç®å½ç»æ:

ååï¼åè¯ï¼

è¯æ§æ æ³¨

å½åå®ä½è¯å«

å¥æ³åæ

WordNetå¯ä»¥è¢«çä½æ¯ä¸ä¸ªåä¹è¯è¯å¸

è¯å¹²æåï¼stemmingï¼ä¸è¯å½¢è¿åï¼lemmatizationï¼

https://www.biaodianfu.com/nltk.html/amp

TensorFlow 2.0å¦ä¹ ç½å

https://github.com/lyhue1991/eat_tensorflow2_in_30_days

3.èªç¶è¯è¨å¤ç

æ¯æçæ¬

Version	Supported
3.6.x	:white_check_mark:
2.7.x	:x:

å¦ä¹ è¿ç¨ä¸-åå¿å¤æçååï¼ï¼ï¼

èªä»å¦ä¹ NLPä»¥åï¼æåç°å½åä¸å½å¤çå¸ååºå«:
1. å¯¹èµæºçæåº¦æ¯å®å¨ç¸åç:
  1) å½å: å°±å¥½åä¸ºäºåæ°ï¼ä¸¾åå·¥ä½è£é¼çä¼è®®ï¼å°±æ¯æ²¡æå¹²è´§ï¼å¨é¨é½æ¯è±¡å¾æ§çPPTä»ç»ï¼ä¸æ¯éå¯¹å¨åçåä½
  2ï¼å½å¤: å°±å¥½åæ¯ä¸ºäºæ¨å¨nlpè¿æ¥ä¸æ ·ï¼åäº«èåç§å¹²è´§èµæåå·ä½çå®ç°ãï¼ç¹å«æ¯: pythonèªç¶è¯è¨å¤çï¼
2. è®ºæçå®ç°: 
  1) åç§é«å¤§ä¸çè®ºæå®ç°ï¼å´è¿æ¯æ²¡çå°ä¸ä¸ªåæ ·çGitHubé¡¹ç®ï¼ï¼å¯è½æçæç´¢è½åå·®äºç¹ï¼ä¸ç´æ²¡æ¾å°ï¼
  2ï¼å½å¤å°±ä¸ä¸¾ä¾äºï¼æçä¸æï¼
3. å¼æºçæ¡æ¶
  1ï¼å½å¤çå¼æºæ¡æ¶:  tensorflow/pytorch ææ¡£+æç¨+è§é¢ï¼å®æ¹æä¾ï¼
  2) å½åçå¼æºæ¡æ¶: é¢é¢ï¼è¿çä¸¾ä¾ä¸åºæ¥ï¼ä½æ¯çé¼å¹å¾ä¸æ¯å½å¤å·®ï¼ï¼MXNetè½ç¶æä¼å¤å½äººåä¸å¼åï¼ä½ä¸è½ç®æ¯å½åå¼æºæ¡æ¶ãåºäºMXNetçå¨æå¦æ·±åº¦å¦ä¹ (http://zh.d2l.ai & https://discuss.gluon.ai/t/topic/753)ä¸ææç¨,å·²ç»ç±æ²ç¥(ææ²)ä»¥åé¿æ¯é¡¿Â·å¼ è®²æå½å¶ï¼å¬å¼åå¸(ææ¡£+ç¬¬ä¸å£æç¨+è§é¢ï¼ã)
æ¯ä¸æ¬¡æ·±å¥é½è¦å»ç¿»å¢ï¼æ¯ä¸æ¬¡æ·±å¥é½è¦Googleï¼æ¯ä¸æ¬¡ççå½åçè¯´: åå·¥å¤§ãè®¯é£ãä¸ç§å¤§ãç¾åº¦ãé¿éå¤çé¼ï¼ä½æ¯èµæè¿æ¯å¾å½å¤å»æ¾ï¼
ææ¶åççæºæ¨çï¼ççæç¹ç§ä¸èµ·èªå·±å½åçææ¯ç¯å¢ï¼

å½ç¶è°¢è°¢å½åå¾å¤åå®¢å¤§ä½¬ï¼ç¹å«æ¯ä¸äºå¥é¨çDemoååºæ¬æ¦å¿µããæ·±å¥çæ°´å¹³æéï¼æ²¡çæã

ãå¥é¨é¡»ç¥ãå¿é¡»äºè§£: https://github.com/apachecn/AiLearning/tree/master/nlp
ãå¥é¨æç¨ãå¼ºçæ¨è: PyTorch èªç¶è¯è¨å¤ç: https://github.com/apachecn/NLP-with-PyTorch
Python èªç¶è¯è¨å¤ç ç¬¬äºç: https://usyiyi.github.io/nlp-py-2e-zh
æ¨èä¸ä¸ªliuhuanyongå¤§ä½¬æ´ççnlpå¨é¢ç¥è¯ä½ç³»: https://liuhuanyong.github.io
å¼æº - è¯åéåºéå:

1.ä½¿ç¨åºæ¯ ï¼ç¾åº¦å¬å¼è¯¾ï¼

ç¬¬ä¸é¨å å¥é¨ä»ç»

1.) èªç¶è¯è¨å¤çå¥é¨ä»ç»

ç¬¬äºé¨å æºå¨ç¿»è¯

2.) æºå¨ç¿»è¯

ç¬¬ä¸é¨å ç¯ç« åæ

ç¬¬åé¨å UNIT-è¯è¨çè§£ä¸äº¤äºææ¯

4.) UNIT-è¯è¨çè§£ä¸äº¤äºææ¯

åºç¨é¢å

ä¸æåè¯:

æå»ºDAGå¾
å¨æè§åæ¥æ¾ï¼ç»¼åæ£ååï¼æ£åå æååè¾åºï¼æ±å¾DAGæå¤§æ¦çè·¯å¾
ä½¿ç¨äºSBMEè¯æè®ç»äºä¸å¥ HMM + Viterbi æ¨¡åï¼è§£å³æªç»å½è¯é®é¢

1.ææ¬åç±»ï¼Text Classificationï¼

ä¸é¢æ¯ä¸äºå¾å¥½çåå¦èææ¬åç±»æ°æ®éã

è·¯éç¤¾Newswireä¸»é¢åç±»ï¼è·¯éç¤¾-21578ï¼ã1987å¹´è·¯éç¤¾åºç°çä¸ç³»åæ°é»æä»¶ï¼æç±»å«ç¼å¶ç´¢å¼ãå¦è§RCV1ï¼RCV2åTRC2ã
IMDBçµå½±è¯è®ºææåç±»ï¼æ¯å¦ç¦ï¼ãæ¥èªç½ç«imdb.comçä¸ç³»åçµå½±è¯è®ºåå¶ç§¯æææ¶æçæç»ªã
æ°é»ç»çµå½±è¯è®ºææåç±»ï¼åº·å¥å°ï¼ãæ¥èªç½ç«imdb.comçä¸ç³»åçµå½±è¯è®ºåå¶ç§¯æææ¶æçæç»ªã

æå³æ´å¤ä¿¡æ¯ï¼è¯·åéå¸å: åæ ç¾ææ¬åç±»çæ°æ®éã

ææåæ

æ¯èµå°å: https://www.kaggle.com/c/word2vec-nlp-tutorial

æ¹æ¡ä¸(0.86): WordCount + æ´ç´ Bayes
æ¹æ¡äº(0.94): LDA + åç±»æ¨¡åï¼knn/å³çæ /é»è¾åå½/svm/xgboost/éæºæ£®æï¼
- a) å³çæ ææä¸æ¯å¾å¥½ï¼è¿ç§è¿ç»ç¹å¾ä¸å¤ªéåç
- b) éè¿åæ°è°æ´ 200 ä¸ªtopicï¼ä¿¡æ¯éä¿åææè¾ä¼ï¼è®¡ç®ä¸»é¢ï¼
æ¹æ¡ä¸(0.72): word2vec + CNN
- è¯´å®è¯: æ²¡æä¸ä¸ªå¥½çæºå¨ï¼æ¯è°ä¸åºæ¥ä¸ä¸ªå¥½çç»æ (: é

éè¿AUC æ¥è¯ä¼°æ¨¡åçææ

2.è¯è¨æ¨¡åï¼Language Modelingï¼

å®æ¯è¯é³è¯å«åæºå¨ç¿»è¯çä»»å¡ä¸çåç½®ä»»å¡ã

ä¸é¢æ¯ä¸äºå¾å¥½çåå¦èè¯è¨å»ºæ¨¡æ°æ®éã

å¤è¾å ¡é¡¹ç®ï¼ä¸ç³»ååè´¹ä¹¦ç±ï¼å¯ä»¥ç¨çº¯ææ¬æ£ç´¢åç§è¯è¨ã

æ°è¯åç°

ä¸æåè¯æ°è¯åç°
https://github.com/zhanzecheng/Chinese_segment_augment

å¥åç¸ä¼¼åº¦è¯å«

é¡¹ç®å°å: https://www.kaggle.com/c/quora-question-pairs
è§£å³æ¹æ¡: word2vec + Bi-GRU

ææ¬çº é

bi-gram + levenshtein

3.å¾ååå¹ï¼Image Captioningï¼

mageåå¹æ¯ä¸ºç»å®å¾åçæææ¬æè¿°çä»»å¡ã

ä¸é¢æ¯ä¸äºå¾å¥½çåå¦èå¾ååå¹æ°æ®éã

ä¸ä¸æä¸çå¬å±å¯¹è±¡ï¼COCOï¼ãåå«è¶è¿12ä¸å¼ å¸¦æè¿°çå¾åçéå
Flickr 8Kãä»flickr.comè·åç8åä¸ªæè¿°å¾åçéåã
Flickr 30Kãä»flickr.comè·åç3ä¸ä¸ªæè¿°å¾åçéåã æ¬²äºè§£æ´å¤ï¼è¯·çå¸å:

æ¢ç´¢å¾ååå¹æ°æ®éï¼2016å¹´

4.æºå¨ç¿»è¯ï¼Machine Translationï¼

æºå¨ç¿»è¯æ¯å°ææ¬ä»ä¸ç§è¯è¨ç¿»è¯æå¦ä¸ç§è¯è¨çä»»å¡ã

ä¸é¢æ¯ä¸äºå¾å¥½çåå¦èæºå¨ç¿»è¯æ°æ®éã

å æ¿å¤§ç¬¬36å±è®®ä¼çåè°å½ä¼è®®åãæå¯¹çè±è¯åæ³è¯å¥åã
æ¬§æ´²è®®ä¼è¯è®¼å¹³è¡è¯æåº1996-2011ãå¥åå¯¹ä¸å¥æ¬§æ´²è¯è¨ã æå¤§éæ åæ°æ®éç¨äºå¹´åº¦æºå¨ç¿»è¯ææ; çå°:

ç»è®¡æºå¨ç¿»è¯

æºå¨ç¿»è¯

Encoder + Decoder(Attention)
åèæ¡ä¾: http://pytorch.apachecn.org/cn/tutorials/intermediate/seq2seq_translation_tutorial.html

5.é®çç³»ç»ï¼Question Answeringï¼

ä¸é¢æ¯ä¸äºå¾å¥½çåå¦èé®é¢åçæ°æ®éã

æ¯å¦ç¦é®é¢åçæ°æ®éï¼SQuADï¼ãåçæå³ç»´åºç¾ç§æç« çé®é¢ã
Deepmindé®é¢åçè¯æåºãä»æ¯æ¥é®æ¥åçæå³æ°é»æç« çé®é¢ã

æ°æ®é: æå¦ä½è·å¾é®çç½ç«çè¯æåºï¼å¦QuoraæYahoo AnswersæStack Overflowæ¥åæçæ¡è´¨éï¼

6.è¯é³è¯å«ï¼Speech Recognitionï¼

ä¸é¢æ¯ä¸äºå¾å¥½çåå¦èè¯é³è¯å«æ°æ®éã

TIMITå£°å¦ - è¯é³è¿ç»è¯é³è¯æåºãä¸æ¯åè´¹çï¼ä½å å¶å¹¿æ³ä½¿ç¨èä¸å¸ãå£è¯ç¾å½è±è¯åç¸å³çè½¬å½ã
VoxForgeãç¨äºæå»ºç¨äºè¯é³è¯å«çå¼æºæ°æ®åºçé¡¹ç®ã

7.èªå¨ææï¼Document Summarizationï¼

ææ¡£æè¦æ¯åå»ºè¾å¤§ææ¡£çç®çææä¹æè¿°çä»»å¡ã

ä¸é¢æ¯ä¸äºå¾å¥½çåå¦èææ¡£æè¦æ°æ®éã

æ³å¾æ¡ä¾æ¥åæ°æ®éãæ¶éäº4000ä»½æ³å¾æ¡ä»¶åå¶æè¦ã
TIPSTERææ¬æè¦è¯ä¼°ä¼è®®è¯æåºãæ¶éäºè¿200ä»½æä»¶åå¶æè¦ã
è±è¯æ°é»ææ¬çAQUAINTè¯æåºãä¸æ¯åè´¹çï¼èæ¯å¹¿æ³ä½¿ç¨çãæ°é»æç« çè¯æåºã æ¬²äºè§£æ´å¤ä¿¡æ¯:

å½åå®ä½è¯å«

Bi-LSTM CRF
åèæ¡ä¾: http://pytorch.apachecn.org/cn/tutorials/beginner/nlp/advanced_tutorial.html
CRFæ¨èææ¡£: https://www.jianshu.com/p/55755fc649b1

ææ¬æè¦

æ½åå¼
word2vec + textrank
word2vecæ¨èææ¡£: https://www.zhihu.com/question/44832436/answer/266068967
textrankæ¨èææ¡£: https://blog.csdn.net/BaiHuaXiu123/article/details/77847232

Graphå¾è®¡ç®ãæ¢æ¢æ´æ°ã

æ°æ®é: https://github.com/apachecn/data/tree/master/graph
å¦ä¹ èµæ: spark graphXå®æ.pdf ãæä»¶å¤ªå¤§ä¸æ¹ä¾¿æä¾ï¼èªå·±ç¾åº¦ã

ç¥è¯å¾è°±

ç¥è¯å¾è°±ï¼æåªè®¤ SimmerChan: ãç¥è¯å¾è°±-ç»AIè£ä¸ªå¤§èã
è¯´å®è¯ï¼ææ¯çè¿åä¸»èå¥åçåå®¢é¿å¤§çï¼åçççæ¯æ·±å¥æµåºãæå¾åæ¬¢ï¼æä»¥å°±åäº«ç»å¤§å®¶ï¼å¸æä½ ä»¬ä¹åæ¬¢ã

è¿ä¸æ¥éè¯»

å¦ææ¨å¸ææ´æ·±å¥ï¼æ¬èæä¾äºå¶ä»æ°æ®éåè¡¨ã

åè

è´è°¢

èµå©æä»¬

Top Related Projects

Convert designs to code with AI

Introducing Visual Copilot: A new AI model to turn Figma designs to high quality code using your components.

Try Visual Copilot

Top Related Projects

Quick Overview

Pros

Cons

Code Examples

Competitor Comparisons

Pros of ML-For-Beginners

Cons of ML-For-Beginners

Code Comparison

Summary

Pros of handson-ml

Cons of handson-ml

Code Comparison

Summary

Pros of python-machine-learning-book

Cons of python-machine-learning-book

Code Comparison

Pros of TensorFlow-Examples

Cons of TensorFlow-Examples

Code Comparison

Pros of ML-From-Scratch

Cons of ML-From-Scratch

Code Comparison

Pros of ML-YouTube-Courses

Cons of ML-YouTube-Courses

Code Comparison

Convert designs to code with AI

README

è·¯çº¿å¾

1.æºå¨å­¦ä¹ - åºç¡

åºæ¬ä»ç»

å­¦ä¹ ææ¡£

ç½ç«è§é¢

2.æ·±åº¦å­¦ä¹

å ¥é¨åºç¡

Pytorch - æç¨

TensorFlow 2.0 - æç¨

3.èªç¶è¯­è¨å¤ç

1.ä½¿ç¨åºæ¯ ï¼ç¾åº¦å ¬å¼è¯¾ï¼

åºç¨é¢å

ä¸­æåè¯:

1.ææ¬åç±»ï¼Text Classificationï¼

2.è¯­è¨æ¨¡åï¼Language Modelingï¼

3.å¾åå­å¹ï¼Image Captioningï¼

4.æºå¨ç¿»è¯ï¼Machine Translationï¼

5.é®ç­ç³»ç»ï¼Question Answeringï¼

6.è¯­é³è¯å«ï¼Speech Recognitionï¼

7.èªå¨ææï¼Document Summarizationï¼

Graphå¾è®¡ç®ãæ ¢æ ¢æ´æ°ã

ç¥è¯å¾è°±

è¿ä¸æ­¥é è¯»

åè

è´è°¢

èµå©æä»¬

Top Related Projects

Convert designs to code with AI

è·¯çº¿å¾

1.æºå¨å¦ä¹ - åºç¡

åºæ¬ä»ç»

å¦ä¹ ææ¡£

ç½ç«è§é¢

2.æ·±åº¦å¦ä¹

å¥é¨åºç¡

Pytorch - æç¨

TensorFlow 2.0 - æç¨

3.èªç¶è¯è¨å¤ç

1.ä½¿ç¨åºæ¯ ï¼ç¾åº¦å¬å¼è¯¾ï¼

åºç¨é¢å

ä¸æåè¯:

1.ææ¬åç±»ï¼Text Classificationï¼

2.è¯è¨æ¨¡åï¼Language Modelingï¼

3.å¾ååå¹ï¼Image Captioningï¼

4.æºå¨ç¿»è¯ï¼Machine Translationï¼

5.é®çç³»ç»ï¼Question Answeringï¼

6.è¯é³è¯å«ï¼Speech Recognitionï¼

7.èªå¨ææï¼Document Summarizationï¼

Graphå¾è®¡ç®ãæ¢æ¢æ´æ°ã

ç¥è¯å¾è°±

è¿ä¸æ¥éè¯»

åè

è´è°¢

èµå©æä»¬