Artificial Neural Networks (ANN), also known simply as neural networks, are mathematical models that closely resemble the characteristics of biological neural networks among many machine learning algorithms. ANN simulates the structure and function of biological neural networks (the brain) and is composed of numerous nodes (also called “neurons” or “units”) that are interconnected, which can be used to model the complex relationships between data. Artificial Neural Networks (ANN) are a set of multilayer perceptrons/neural units. ANN is also referred to as feedforward neural networks because the input is only processed in a forward direction:ANN consists of three layers: the input layer, hidden layer, and output layer. The input layer accepts inputs, the hidden layer processes the inputs, and the output layer generates results. Essentially, each layer attempts to learn certain weights.Artificial Neural Networks can learn any nonlinear function. Therefore, these networks are widely known as Universal Function Approximators. ANN has the ability to learn the weights that map any input to an output. One of the main reasons for universal approximation is the activation function. The activation function introduces nonlinearity into the network. This helps the network learn any complex relationship between inputs and outputs. As you can see, the output of each neuron is the activation of the weighted sum of the inputs. Without an activation function, the network can only learn linear functions and cannot learn complex relationships. Therefore, the activation function is the driving force behind artificial neural networks.
2 Algorithm Principles
The structure of the artificial neural network is shown in the figure below:
Round nodes represent artificial neurons: In an artificial neural network, each round node represents an artificial neuron. These neurons interact with each other through specific connections, simulating the working principle of biological neural networks.Connections and Signal Transmission: Arrows indicate the connections from the output of one neuron to the input of another neuron. Through these connections, signals can be transmitted within the network, from one artificial neuron to another.Weights and Activation Functions: Each node represents a specific output function, called the activation function. Each connection between two nodes has an associated weight value that indicates the degree of influence the previous neuron has on the next neuron.Network Output: The output of the network varies based on the network’s connections, weight values, and activation functions.By adjusting these parameters, artificial neural networks can learn and adapt to different input patterns, producing the expected output results.
3 Algorithm Applications
In recent years, artificial neural networks have provided a data-driven solution that can adapt to high-dimensional complex functions, thus having widespread applications in the modeling of traditional Chinese medicine diagnosis and treatment systems. For example, in the research of constructing complex giant systems of traditional Chinese medicine diagnosis and treatment, researchers abstracted the hierarchical elements, contact methods, and system functions based on the prototype of the complex giant system of traditional Chinese medicine diagnosis and treatment, constructing a logical model of the complex giant system of traditional Chinese medicine diagnosis and treatment and elucidating its complexity characteristics. The nonlinearity and complexity characteristics of symptoms add considerable difficulty to the mathematical modeling of symptoms, while the algorithmic mechanism of artificial neural networks shares similarities with the operational model of complex giant systems of traditional Chinese medicine diagnosis and treatment. Therefore, it can effectively simulate the complexity characteristics and functions of the system. This study points out that the core of constructing the complex giant system of traditional Chinese medicine diagnosis and treatmentfollows the methodological principles of complexity science, providing a theoretical basis for the application of artificial neural networks in the modeling of traditional Chinese medicine diagnosis and treatment systems, and laying the foundation for the interpretability research of specific models.
4 Summary
ANN can simulate the hierarchical elements, contact methods, functions, and complexity characteristics of the complex giant system of traditional Chinese medicine diagnosis and treatment without needing to know the internal structure of the system. Therefore, it is an important method for modeling the complex giant system of traditional Chinese medicine diagnosis and treatment, providing a theoretical basis for the application of ANN in this field.At the same time, the weak interpretability caused by the black-box structure of ANN is a significant obstacle to its development and application. In practical modeling processes, there are challenges in selecting suitable “hidden layer numbers” and “node numbers” during model training, along with long training times, high spatiotemporal complexity, and low accuracy of results. Therefore, exploring the interpretability of ANN models is a key research direction for the next step.References:[1] Sun Zhaoyang, Lu Guijiao, Guo Yi, et al. Principles of Artificial Neural Network Modeling for Complex Giant Systems in Traditional Chinese Medicine Diagnosis and Treatment [J]. Chinese Journal of Traditional Chinese Medicine, 2022, 37(10):5841-5844.[2] Zhihu Column. “CNN vs. RNN vs. ANN – A Brief Analysis of Three Neural Networks in Deep Learning”. Accessed on April 23, 2024.[3] “Summary of the Algorithm of Artificial Neural Network ANN – ANN Algorithm – CSDN Blog”. Accessed on April 23, 2024.[4] “Neural Network Algorithm – Understanding ANN (Artificial Neural Network) in One Article – ANN Neural Network – CSDN Blog”. Accessed on April 23, 2024.Recommended Reading:Markov Prediction Model – “The Future is Independent of the Past, Only Based on the Present”Transformer Model – The Organic Combination of Attention Mechanism and Neural NetworksThe Era of Large Language Models: Efficiency and Precision Coexist, Opportunities and Challenges Exist
Ancient and Modern Medical Case Cloud Platform
Providing Retrieval Services for Over 500,000 Ancient and Modern Medical Cases
Supports Manual, Voice, OCR, and Batch Structured Entry of Medical Cases
Designs Nine Major Analysis Modules, Close to Clinical Practical Needs
Supports Collaborative Analysis of Massive Medical Cases and Personal Medical Cases
EDC Traditional Chinese Medicine Research Case Collection System
Supports Multi-Center, Central Randomization, Data Entry,
SDV, Audit Trail, SMS Reminders, Data Statistics
Analysis, Patient Diary Cards, Mobile Entry, etc.
Supports Customized Form Design
Users can log in at: https://www.yiankb.com/edc
Institute of Traditional Chinese Medicine Information Research, Chinese Academy of Traditional Chinese Medicine
Intelligent R&D Center for Traditional Chinese Medicine Health
