"Despite the enormous success of deep learning, relatively little is understood theoretically about why these techniques are so successful at feature learning and compression." (Pankaj Mehta & David J Schwab, "An exact mapping between the Variational Renormalization Group and Deep Learning", 2014)
"Deep learning is about using a stacked hierarchy of feature detectors. [...] we use pattern detectors and we build them into networks that are arranged in hundreds of layers and then we adjust the links between these layers, usually using some kind of gradient descent."
"The power of deep learning models comes from their ability to classify or predict nonlinear data using a modest number of parallel nonlinear steps4. A deep learning model learns the input data features hierarchy all the way from raw data input to the actual classification of the data. Each layer extracts features from the output of the previous layer." (N D Lewis, "Deep Learning Made Easy with R: A Gentle Introduction for Data Science", 2016)
"Although deep learning systems share some similarities with machine learning systems, certain characteristics make them sufficiently distinct. For example, conventional machine learning systems tend to be simpler and have fewer options for training. DL systems are noticeably more sophisticated; they each have a set of training algorithms, along with several parameters regarding the systems’ architecture. This is one of the reasons we consider them a distinct framework in data science." (Yunus E Bulut & Zacharias Voulgaris, "AI for Data Science: Artificial Intelligence Frameworks and Functionality for Deep Learning, Optimization, and Beyond", 2018)
"Deep learning broadly describes the large family of neural network architectures that contain multiple, interacting hidden layers." (Benjamin Bengfort et al, Applied Text Analysis with Python, 2018)
"Deep learning has instead given us machines with truly impressive abilities but no intelligence. The difference is profound and lies in the absence of a model of reality." (Judea Pearl, "The Book of Why: The New Science of Cause and Effect", 2018)
"DL systems also tend to be more autonomous than their machine counterparts. To some extent, DL systems can do their own feature engineering. More conventional systems tend to require more fine-tuning of the feature-set, and sometimes require dimensionality reduction to provide any decent results. In addition, the generalization of conventional ML systems when provided with additional data generally don’t improve as much as DL systems. This is also one of the key characteristics that makes DL systems a preferable option when big data is involved." (Yunus E Bulut & Zacharias Voulgaris, "AI for Data Science: Artificial Intelligence Frameworks and Functionality for Deep Learning, Optimization, and Beyond", 2018)
"[…] deep learning has succeeded primarily by showing that certain questions or tasks we thought were difficult are in fact not. It has not addressed the truly difficult questions that continue to prevent us from achieving humanlike AI." (Judea Pearl & Dana Mackenzie, "The Book of Why: The new science of cause and effect", 2018)
"In essence, deep learning models are just chains of functions, which means that many deep learning libraries tend to have a functional or verbose, declarative style." (Benjamin Bengfort et al, Applied Text Analysis with Python, 2018)
"The second big myth of data science is that every data science project needs big data and needs to use deep learning. In general, having more data helps, but having the right data is the more important requirement" (John D Kelleher & Brendan Tierney, "Data Science", 2018)
"Deep learning approaches the problem of representation learning by introducing representations that are expressed in terms of other, simpler representations. In deep learning, the machine builds multiple levels of increasing complexity over the underlying simpler concepts." (Alessandro Negro, "Graph-Powered Machine Learning", 2021)
"Graphs, with multiple node types and different types of relationships, are far from being a Euclidean space. This task is where graph neural networks (GNNs) comes in. GNNs are deep learning-based methods that operate on a graph domain to perform complex tasks such as node classification (the bot example), link prediction (the disease example), and so on. Due to its convincing performance, GNN has become a widely applied graph analysis method." (Alessandro Negro, "Graph-Powered Machine Learning", 2021)
"People who assume that extensions of modern machine learning methods like deep learning will somehow 'train up', or learn to be intelligent like humans, do not understand the fundamental limitations that are already known. Admitting the necessity of supplying a bias to learning systems is tantamount to Turing’s observing that insights about mathematics must be supplied by human minds from outside formal methods, since machine learning bias is determined, prior to learning, by human designers."
"Deep learning algorithms are exceptionally adept at identifying anomalies within large datasets, making them an indispensable tool for detecting fraudulent transactions and financial irregularities. By learning from historical transaction data, these models can pinpoint patterns and behaviors indicative of fraudulent activities with remarkable accuracy. This ability not only aids in safeguarding assets but also ensures compliance with increasingly stringent regulatory standards aimed at preventing financial fraud and misconduct." (Hayden Van Der Post, "Neural Network: Mastering the Art of Algorithmic Trading", 2024)
"Deep learning is an advanced subset of machine learning, distinguished by its ability to process data through layers of neural networks, each layer abstracting information from the one preceding it. This hierarchical approach enables the model to handle complex, high-dimensional data, learning features and patterns at multiple levels of abstraction. [...] Traditional neural networks, with their shallower architectures, often struggle with the nuances of financial data, limited by their capacity to extrapolate and interpret intricate patterns. Deep learning, however, with its deeper, more sophisticated networks, can navigate these complexities, offering nuanced insights into market dynamics." (Hayden Van Der Post, "Neural Network: Mastering the Art of Algorithmic Trading", 2024)
"Language modeling is a subfield of NLP that involves the creation of statistical/deep learning models for predicting the likelihood of a sequence of tokens in a specified vocabulary (a limited and known set of tokens). There are generally two kinds of language modeling tasks out there: autoencoding tasks and autoregressive tasks." (Sinan Ozdemir, "Quick Start Guide to Large Language Models: Strategies and Best Practices for Using ChatGPT and Other LLMs", 2024)
"LLMs are advanced natural language processing (NLP) deep learning models. They specialize in both processing contextual language at scale and predicting the likelihood of a sequence of tokens in a specific language. The smallest units of semantic meaning, tokens can be words or sub-words and act as the key inputs for an LLM. LLMs can be categorized as autoregressive, autoencoding, or a combination of both. Their defining feature is their substantial size, which enables them to execute complex language tasks like text generation and classification, with high precision and potentially minimal fine-tuning." (Sinan Ozdemir, "Quick Start Guide to Large Language Models: Strategies and Best Practices for Using ChatGPT and Other LLMs", 2024)
"[...] LLMs are pre-trained on large corpora and sometimes fine-tuned on smaller datasets for specific tasks. Recall that one of the factors behind the Transformer’s effectiveness as a language model is that it is highly parallelizable, allowing for faster training and efficient processing of text. What really sets the Transformer apart from other deep learning architectures is its ability to capture long-range dependencies and relationships between tokens using attention. In other words, attention is a crucial component of Transformer-based LLMs, and it enables them to effectively retain information between training loops and tasks (i.e., transfer learning), while being able to process lengthy swatches of text with ease." (Sinan Ozdemir, "Quick Start Guide to Large Language Models: Strategies and Best Practices for Using ChatGPT and Other LLMs", 2024)
"As with many other deep learning-based approaches, another major challenge is in interpretability. While knowledge graphs provide a structured and transparent way to store relationships, LLMs operate as a black box, making it difficult to understand how specific outputs are generated. [...] Data alignment is also a key issue, as structured knowledge graphs and unstructured text data must be carefully preprocessed to ensure consistency. Differences in data formats, ontology mismatches, and information redundancy can create inefficiencies when integrating these two paradigms. Developing robust pipelines that seamlessly connect graph-based insights with LLM-generated text remains an open challenge." (Aldo Marzullo et al, "Graph Machine Learning" 2nd Ed., 2025)
"The same difficulties that characterize training deep feedforward networks also apply to RNNs; gradients tend to die out over long distances using traditional activation functions (or explode if the gradients become greater than 1). However, unlike feedforward networks, RNNs aren’t trained with traditional backpropagation, but rather a variant known as Backpropagation through Time (BPTT): the network is unrolled, as before, and backpropagation is used, averaging over errors at each time point (since an 'output', the hidden state, occurs at each step). Also, in the case of RNNs, we run into the problem that the network has a very short memory; it only incorporates information from the most recent unit before the current one and has trouble maintaining long-range context. For applications such as translation, this is clearly a problem, as the interpretation of a word at the end of a sentence may depend on terms near the beginning, not just those directly preceding it." (Joseph Babcock & Raghav Bali, "Generative AI with Python and PyTorch" 2nd. Ed., 2025)
No comments:
Post a Comment