اشترك بالحزمة الذهبية واحصل على وصول غير محدود شمرا أكاديميا
تسجيل مستخدم جديدDeep Reinforcement Learning Models Predict Visual Responses in the Brain: A Preliminary Result
137
0
0.0
(
0
)
اسأل ChatGPT حول البحث
ﻻ يوجد ملخص باللغة العربية
Supervised deep convolutional neural networks (DCNNs) are currently one of the best computational models that can explain how the primate ventral visual stream solves object recognition. However, embodied cognition has not been considered in the existing visual processing models. From the ecological standpoint, humans learn to recognize objects by interacting with them, allowing better classification, specialization, and generalization. Here, we ask if computational models under the embodied learning framework can explain mechanisms underlying object recognition in the primate visual system better than the existing supervised models? To address this question, we use reinforcement learning to train neural network models to play a 3D computer game and we find that these reinforcement learning models achieve neural response prediction accuracy scores in the early visual areas (e.g., V1 and V2) in the levels that are comparable to those accomplished by the supervised neural network model. In contrast, the supervised neural network models yield better neural response predictions in the higher visual areas, compared to the reinforcement learning models. Our preliminary results suggest the future direction of visual neuroscience in which deep reinforcement learning should be included to fill the missing embodiment concept.
قيم البحث
اقرأ أيضاً
Can machine learning help us make better decisions about a changing planet? In this paper, we illustrate and discuss the potential of a promising corner of machine learning known as _reinforcement learning_ (RL) to help tackle the most challenging co
nservation decision problems. RL is uniquely well suited to conservation and global change challenges for three reasons: (1) RL explicitly focuses on designing an agent who _interacts_ with an environment which is dynamic and uncertain, (2) RL approaches do not require massive amounts of data, (3) RL approaches would utilize rather than replace existing models, simulations, and the knowledge they contain. We provide a conceptual and technical introduction to RL and its relevance to ecological and conservation challenges, including examples of a problem in setting fisheries quotas and in managing ecological tipping points. Four appendices with annotated code provide a tangible introduction to researchers looking to adopt, evaluate, or extend these approaches.
The feasibility of wave function collapse in the human brain has been the subject of vigorous scientific debates since the advent of quantum theory. Scientists like Von Neumann, London, Bauer and Wigner (initially) believed that wave function collaps
e occurs in the brain or is caused by the mind of the observer. It is a legitimate question to ask how human brain can receive subtle external visual quantum information intact when it must pass through very noisy and complex pathways from the eye to the brain? There are several approaches to investigate information processing in the brain, each of which presents a different set of conclusions. Penrose and Hameroff have hypothesized that there is quantum information processing inside the human brain whose material substrate involves microtubules and consciousness is the result of a collective wavefunction collapse occurring in these structures. Conversely, Tegmark stated that owing to thermal decoherence there cannot be any quantum processing in neurons of the brain and processing in the brain must be classical for cognitive processes. However, Rosa and Faber presented an argument for a middle way which shows that none of the previous authors are completely right and despite the presence of decoherence, it is still possible to consider the brain to be a quantum system. Additionally, Thaheld, has concluded that quantum states of photons do collapse in the human eye and there is no possibility for collapse of visual quantum states in the brain and thus there is no possibility for the quantum state reduction in the brain. In this paper we conclude that if we accept the main essence of the above approaches taken together, each of them can provide a different part of a teleportation mechanism.
Brain Electroencephalography (EEG) classification is widely applied to analyze cerebral diseases in recent years. Unfortunately, invalid/noisy EEGs degrade the diagnosis performance and most previously developed methods ignore the necessity of EEG se
lection for classification. To this end, this paper proposes a novel maximum weight clique-based EEG selection approach, named mwcEEGs, to map EEG selection to searching maximum similarity-weighted cliques from an improved Fr{e}chet distance-weighted undirected EEG graph simultaneously considering edge weights and vertex weights. Our mwcEEGs improves the classification performance by selecting intra-clique pairwise similar and inter-clique discriminative EEGs with similarity threshold $delta$. Experimental results demonstrate the algorithm effectiveness compared with the state-of-the-art time series selection algorithms on real-world EEG datasets.
This paper describes a scalable active learning pipeline prototype for large-scale brain mapping that leverages high performance computing power. It enables high-throughput evaluation of algorithm results, which, after human review, are used for iter
ative machine learning model training. Image processing and machine learning are performed in a batch layer. Benchmark testing of image processing using pMATLAB shows that a 100$times$ increase in throughput (10,000%) can be achieved while total processing time only increases by 9% on Xeon-G6 CPUs and by 22% on Xeon-E5 CPUs, indicating robust scalability. The images and algorithm results are provided through a serving layer to a browser-based user interface for interactive review. This pipeline has the potential to greatly reduce the manual annotation burden and improve the overall performance of machine learning-based brain mapping.
Reinforcement Learning (RL) is a key technique to address sequential decision-making problems and is crucial to realize advanced artificial intelligence. Recent years have witnessed remarkable progress in RL by virtue of the fast development of deep
neural networks. Along with the promising prospects of RL in numerous domains, such as robotics and game-playing, transfer learning has arisen as an important technique to tackle various challenges faced by RL, by transferring knowledge from external expertise to accelerate the learning process. In this survey, we systematically investigate the recent progress of transfer learning approaches in the context of deep reinforcement learning. Specifically, we provide a framework for categorizing the state-of-the-art transfer learning approaches, under which we analyze their goals, methodologies, compatible RL backbones, and practical applications. We also draw connections between transfer learning and other relevant topics from the RL perspective and explore their potential challenges as well as open questions that await future research progress.
الأسئلة المقترحة
سجل دخول لتتمكن من نشر تعليقات
التعليقات
جاري جلب التعليقات
سجل دخول لتتمكن من متابعة معايير البحث التي قمت باختيارها
