Artificial intelligence and neuron content combing
- 1. Background
- 2. Preparation basis
- 3. Books
- 4. Community and resources
- 5. Tools and framework
- 6. Learning path
- 7. Related Papers[^1]
-
- 1. Computer Vision Related Surveys
-
- A). Object Detection
- B). Image Segmentation
- C). Medical Image Segmentation
- D). Face Recognition
- E). Biometrics Recognition
- F). Image Super-Resolution
- G). Facial Expression Recognition
- H). Action Recognition
- I). Generative Adversarial Networks (GANs)
- J). Image Captioning
- K). Human Motion Recognition
- L). Neural Rendering
- 2. NLP Related Surveys
-
- A). General Natural Language Processing (General NLP)
- B). Dialogue Systems, Conversational AI (Dialogue Systems, Conversational AI)
- C). Embedding
- D). Natural Language Generation
- E). Text Summarization
- F). Sentiment Analysis
- G). Named Entity Recognition
- H). Answer Selection
- I). NLP based Financial Forecasting
- 3. Healthcare Related Surveys
- 4. Other Areas Surveys
1. Background
Over the past few years, deep learning models have been applied to various tasks/applications in different domains. With the maturity and increase of deep learning in different fields, I decided to organize relevant materials for learning.
2. Preparation basis
Neurons (optional) reading
3. Books
1. "Calculus"
- Function Limit and Continuity
- Derivatives and Differentiation
- indefinite integral
- Definite integral
- Ordinary Differential Equations and Mathematical Models
- Function Integrability and Generalized Integral
- infinite series
- Power series and Fourier series
- Space Analytic Geometry
- Differential Calculus of Multivariate Functions
- Re-integration
- Curve and Surface Integrals
- Vector Field Preliminary
- Higher Order Linear ODEs
2. "Linear Algebra"
- determinant
- system of linear equations
- Matrices and their operations
- Linear Space and Linear Transformation
- Eigenvalues and quadratic forms of matrices
- vector space
- Finite-dimensional vector spaces
- linear map
- polynomial
- Eigenvalues and Eigenvectors
- inner product space
- Operators on Inner Product Spaces
- Operators on Complex Vector Spaces
- Operators on real vector spaces
- trace and determinant
3. "Statistics"
4.《Deep Learning》
- Fundamentals of Applied Mathematics and Machine Learning
- Deep Web and Modern Practice
- deep learning research
4. Community and resources
5. Tools and framework
- torch, torch official document
- tensorflow2+, tensorflow official website tutorial
Preheating
- A Survey of Deep Learning Research
- A Survey on Deep Learning: Algorithms, Techniques, and Applications
6. Learning path
1. Logistic regression, linear regression
y ^ = σ ( w T x + b ) , where σ ( z ) = 1 1 + e − zx ∈ R n (the input of the model, with n features) y ∈ [ 0 , 1 ] (the output of the model, the probability value) w ∈ R n , b ∈ R (model parameters, training content) y ^ = w T x + b (linear regression) \widehat y\;=\;\sigma(w^Tx+b),\; where\;\sigma(z)\;=\;\frac1{1+e^{-z}}\\x\in\mathbb{R}^n (\mathrm{input of the model}, with n\mathrm {features})\\y\in\lbrack0,1\rbrack(\mathrm{output of the model}, \mathrm{probability value})\\w\in\mathbb{R}^n,b\in R( \mathrm{model parameters}, \mathrm{training content})\\\widehat y\;=\;w^Tx+b(\mathrm{linear regression})\\\\\\\\y =s ( wTx+b),whereσ ( z )=1+e−z1x∈Rn (input to the model, withnfeatures)y∈[0,1 ] ( the output of the model , the probability value )w∈Rn,b∈R ( model parameters , training content )y =wTx+b ( linear regression )
2. Loss function
Example
feature n-dimensional data M pieces ∣ abc ⋯ efg ⋯ hij ⋯ ⋮ ⋯ ∣ pass rate 1 pass rate 0 pass rate 1 \\\mathrm{feature}n dimension \\\begin{array}{c}data\\data\ \M\\article\end{array}\;\begin{vmatrix}a&b&c&\cdots\\e&f&g&\cdots\\h&i&j&\cdots\\&\vdots&&\cdots\end{vmatrix}\begin{array}{c} \mathrm{pass rate}\;1\\\mathrm{pass rate}\;0\\\mathrm{pass rate}\;1\\\;\end{array}\\\\\\\\feature n dimensionnumberaccording toMStrip
aehbfi⋮cgj⋯⋯⋯⋯
Pass rate1Pass rate0Pass rate1
设 i ∈ [ 1 , M ] i\in\lbrack1,M\rbrack i∈[1,M], y i ^ \;\widehat{y_i}\; yi Represents probability, Target=1 means qualified, Target=0 means unqualified.
P ( T a r g e t = 1 ∣ x i ) = y i ^ \;P(Target\;=\;1\vert\;x_i\;)=\;\widehat{y_i}\;\\ P(Target=1∣xi)=yi
P ( T a r g e t = 0 ∣ x i ) = 1 − y i ^ \;P(Target\;=\;0\vert\;x_i\;)=1-\;\widehat{y_i}\;\\ P(Target=0∣xi)=1−yi
Loss of function L(w) : L ( w ) = ∏ yi ^ yi ( 1 − yi ^ ) 1 − yi L(w)={\textstyle\prod_{}}\;\widehat{y_i}^{y_i }{(1-\widehat{y_i})}^{1-y_i}L(w)=∏yi
yi(1−yi
)1−yi, among which yi y_iyirepresent the real value.
For the convenience of calculation, the above formula can be obtained after equivalent transformation: L ( w ) = ∑ [ yi ln ( yi ^ ) + ( 1 − yi ) ln ( 1 − yi ^ ) ] L(w)={\ textstyle\sum_{}}\lbrack y_i\ln\left(\widehat{y_i}\right)+(1-y_i\;)\ln\left(1-\widehat{y_i}\right)\rbrackL(w)=∑[yiln(yi
)+(1−yi)ln(1−yi
)]
pending upgrade
7. Related Papers 1
Due to the sheer number of works published in the field of deep learning, below is a comprehensive survey of papers published (or under review) in top journals
1. Computer Vision Related Surveys
Here, we provide a list targeting deep learning applications in computer vision.
A). Object Detection
Deep Learning for General Object Detection: A Survey (IJCV, 2019)
Object Detection Using Deep Learning: A Review (IEEE TNNLS, 2019)
A Review of Deep Learning-Based Object Detection (IEEE Access, 2019)
Recent Advances in Deep Learning for Object Detection (Neurocomputing, 2020)
B). Image Segmentation
Image Segmentation Using Deep Learning: A Survey (IEEE PAMI, 2021) A
Survey of Deep Learning Techniques for Semantic Segmentation of Images and Videos (Applied Soft Computing, 2018)
C). Medical Image Segmentation
Deep Learning Techniques for Medical Image Segmentation: Achievements and Challenges (Journal of Digital Imaging, 2018)
Review: Deep Learning for Medical Image Segmentation Using Multimodal Fusion (Elsevier Array, 2019)
Automatic Breast Ultrasound Image Segmentation: A Survey (Pattern Recognition, 2018)
D). Face Recognition
Deep Face Recognition: A Survey (Arxiv, 2019)
Face Recognition Systems: A Survey (Sensors, 2020)
E). Biometrics Recognition
Biometric Recognition Using Deep Learning (AIRE, 2022)
Deep Learning for Biometric Recognition (ACM CSUR, 2018)
F). Image Super-Resolution
Deep Learning for Image Super-Resolution: A Survey (PAMI, 2020)
Super-Resolution Depth Journey: A Survey (ACM CSUR, 2020)
G). Facial Expression Recognition
Deep Facial Expression Recognition: A Survey (Arxiv, 2019)
H). Action Recognition
Insights into Action Recognition: A Survey (Image and Visual Computing, 2017)
Human Action Recognition and Prediction: A Survey (IJCV, 2022)
I). Generative Adversarial Networks (GANs)
Generative Adversarial Networks: A Survey and Taxonomy (Arxiv, 2019)
Generative Adversarial Networks: A Survey of Private and Secure Applications (ACM CSUR, 2021)
J). Image Captioning
A Comprehensive Survey of Deep Learning for Image Captioning (ACM Computing Surveys, 2019)
K). Human Motion Recognition
L). Neural Rendering
The State of the Art in Neural Rendering (Computer Graphics Forum, 2020)
2. NLP Related Surveys
NLP tasks have made great progress in the past few years with the help of deep neural architectures, such as: attentional LSTM, Transformer, BERT, GPT models, and XL-Net.
A). General Natural Language Processing (General NLP)
Text Classification Based on Deep Learning: A Comprehensive Review (ACM Computing Surveys, 2021)
Latest Trends in Natural Language Processing Based on Deep Learning (IEEE Computational intelligence magazine, 2018)
B). Dialogue Systems, Conversational AI (Dialogue Systems, Conversational AI)
Approaches to Conversational AI (Foundations and Trends in Information Retrieval®, 2019) A
Survey of Dialogue Systems-Focused Natural Language Generation Techniques – Past, Present, and Future Directions (Arxiv, 2018) A
Survey of Dialogue Systems: Recent Advances and New Domain (Acm Sigkdd Explorations Newsletter, 2017)
C). Embedding
Word Embeddings (Arxiv, 2019)
Neural Information Retrieval: A Literature Review (Arxiv, 2016)
D). Natural Language Generation
A Survey of the State of Natural Language Generation Technologies: Core Tasks, Applications, and Evaluation (Journal of Artificial Intelligence Research, 2018)
E). Text Summarization
Neural Abstraction Text Summarization Using Sequence-to-Sequence Models (Arxiv, 2019)
F). Sentiment Analysis
Deep Learning for Sentiment Analysis (Wiley Interdisciplinary Review: Data Mining and Knowledge Discovery, 2018)
Deep Learning for Aspect-Based Sentiment Analysis: A Comparative Review (Expert Systems with Applications, 2019)
Survey of Social Media Sentiment Analysis (Knowledge and Information Systems, 2019)
G). Named Entity Recognition
A Survey of Recent Advances in Named Entity Recognition with Deep Learning Models (Arxiv, 2019)
A Survey of Deep Learning for Named Entity Recognition (Arxiv, 2018)
H). Answer Selection
I). NLP based Financial Forecasting
Natural Language-Based Financial Forecasting: A Survey (Artificial Intelligence Review, 2018)
3. Healthcare Related Surveys
deep learning model
A). Medical Image Analysis
A Survey of Deep Learning in Medical Image Analysis (Medical Image Analysis, 2017)
B). Health-Record Analysis
C). Bioinformatics
A Survey of Data Mining and Deep Learning in Bioinformatics (Journal of Healthcare Systems, 2018)
D). Medical Imaging (MRI - Medical Imaging)
E). Radiotherapy
A Survey of Deep Learning in Radiation Therapy (Biomedical Computing, 2018)
F). Pharmacogenomics
G). Cancer Detection and Diagnosis
Deep Learning for Image-Based Cancer Detection and Diagnosis (Pattern Recognition, 2018)
H). Microscopy Image Analysis
I). Radiology
4. Other Areas Surveys
In recent years, deep learning models have been used in various other fields. Some of the most popular include
A). Reinforcement Learning
Deep Reinforcement Learning: An Overview (Arxiv, 2017)
Deep Reinforcement Learning: A Brief Survey (IEEE Journal of Signal Processing, 2017)
B). Recommender Systems
Deep Learning-Based Recommender Systems: Surveys and New Perspectives (ACM Computing Survey, 2019)
C). Big Data
Big Data Deep Learning Survey (Information Fusion, 2018)
D). Networking
Deep Learning in Mobile and Wireless Networks (IEEE Communications Surveys, 2019)
E). IoT Big Data
Deep learning for IoT big data and streaming analytics: A survey., 2019
F). Anomly Detection
A Survey of Network Anomaly Detection Based on Deep Learning (Cluster Computing, 2017)
G). Remote sensing
H). Mobile Multimedia
I). Graph theory (Graphs)
Deep Learning for Graphs (Arxiv, 2018)
Graph Embedding Techniques, Applications, and Performance (Knowledge-Based Systems, 2018)
J). Multimodal Learning
Deep Multimodal Learning: A Survey of Recent Advances and Trends (IEEE Signal Processing Magazine, 2017)
Deep Multimodal Representation Learning (IEEE Access, 2019)
K). Agriculture
Deep Learning in Agriculture (Computers and electronics in agriculture, 2018)
https://github.com/shervinmin/awesome-deep-learning-survey-papers ↩︎