Download PDFOpen PDF in browser

AGI: Large-Scale Neural Network Modelling

EasyChair Preprint 2200

9 pagesDate: December 19, 2019

Abstract

One of the interesting applications is in creating computers at the AGI ( Artificial General Intelligence ) level equal to human intelligence. Designing AGI is basically creating the human brain on a computer. General intelligence has tons of components and required testing on many functions from image recognition, to the ability to write an essay, to solving Inverse Kinematic problems, etc. In this paper, we present a AGI – a large scale neural network model to achieve General Intelligence involving different components. The AGI model contains three subsystems : (1) EEG based system where Xiang Zhang et all, proposed a novel deep neural network based learning framework that affords perceptive insights into the relationship between the EEG data and brain activities and designed a joint convolutional recurrent neural network that simultaneously learns robust high-level feature presentations through low-dimensional dense embeddings from raw EEG signals. The proposed approach has been to use results of this study as it is and use simulated conditions as true input for our  study; (2) Image system that contains an encoder to convert the input into abstract representations, and a deep image reconstruction which optimizes the output of the decoded images so that it more closely resembles the actual or true images, in combination with a multi-layered convolutional neural network ( CNN ) to simulate the same processes that occur when a human brain perceives images; (3) a LSTM that combines inputs in the forms of both EEG and Image, and predict text symbols associated with images and next images accordingly. In this work, the proposed AGI model illustrates the ability to incrementally learn different functions and form a machine programming loop that enables interactions between EEG signals and Image system, and possibly possess human-like general intelligence.

Keyphrases: Artificial General Intelligence, Artificial Intelligence, Artificial Narrow Intelligence, Artificial Neural Networks, General Intelligence Patterns, Human Intelligence

BibTeX entry
BibTeX does not have the right entry for preprints. This is a hack for producing the correct reference:
@booklet{EasyChair:2200,
  author    = {Poondru Prithvinath Reddy},
  title     = {AGI:  Large-Scale Neural Network Modelling},
  howpublished = {EasyChair Preprint 2200},
  year      = {EasyChair, 2019}}
Download PDFOpen PDF in browser