Reservoirs Learn to Learn
نویسندگان
چکیده
We consider reservoirs in the form of liquid state machines, i.e., recurrently connected networks spiking neurons with randomly chosen weights. So far only weights a linear readout were adapted for specific task. wondered whether performance machines can be improved if recurrent are purpose, rather than randomly. After all, connections brain also not assumed to chosen. Rather, these probably optimized during evolution, development, and prior learning experiences task domains. In order examine benefits choosing within we applied Learning-to-Learn (L2L) paradigm our model: -- hence dynamics machine large family potential tasks, which network might have learn later through modification neurons. found that this two-tiered process substantially improves speed tasks. fact, increases further one does train readouts at relies instead on internal fading memory remembering salient information it could extract from preceding examples current This second type has recently been proposed underlie fast prefrontal cortex motor cortex, is interest explore its models. Since share many properties other types reservoirs, results raise question L2L conveys similar reservoirs.
منابع مشابه
How to Learn/What to Learn
Abst rac t : This paper discusses the kind of in fo rmat ion that must be present in a computer program that models the l i n g u i s t i c development of a c h i l d . A three stage model is presented that character izes the development of a na tu ra l l a n guage parser in a c h i l d of ages one, one and a h a l f , and two. Some data from ch i ld ren of these ages is presented. General prob...
متن کاملLearn to Communicate and Communicate to Learn
Thinking of a computer as an educational tool emphasizes a solitary interaction between the learner and the computer. A tool is something that is applied to an object in order to change it, and this usually implies a single user working on a single object or project. While tools are an important aspect of an educational experience, the tool metaphor can isolate individual learners from each oth...
متن کاملTo Learn or Not to Learn ...
Multiagent systems in which agents interact with each other are now being proposed as a solution to many problems which can be grouped together under the \distributed problem solv-ing" umbrella. For such systems to work properly , it is necessary that agents learn from their environment and adapt their behaviour accordingly. In this paper we present a system which uses a combination of neuro{fu...
متن کاملA model to learn from mistakes (using TOPSIS)
Learning is a new paradigm that is introduced in today's organizations. The purpose of this study is to determine learning contingency model from the mistakes using TOPSIS. Variables were identified using the Delphi technique. The samples of this study were consisted 30 public administrators in Iran. The results show which factors have the greatest impact and which causes the least impact on le...
متن کاملذخیره در منابع من
با ذخیره ی این منبع در منابع من، دسترسی به آن را برای استفاده های بعدی آسان تر کنید
ژورنال
عنوان ژورنال: Natural computing series
سال: 2021
ISSN: ['1619-7127', '2627-6461']
DOI: https://doi.org/10.1007/978-981-13-1687-6_3