(i) may be modified to a continuous model
(ii) can be used for associative memory problems
(iii) nodes of the network have a continuous graded output
(iv) energy of the network increases continuously with time
This solution explains the properties of a discrete Hopfield network based on the provided statements.
This statement is true. The discrete Hopfield network architecture serves as a foundation that can be adapted into a continuous Hopfield network. The continuous version uses differential equations for state transitions instead of discrete updates, allowing for a different range of dynamics and outputs.
This statement is true. One of the primary and most well-known applications of Hopfield networks is in associative memory. They excel at storing multiple patterns and retrieving a stored pattern even when presented with incomplete or noisy input, effectively acting like a form of content-addressable memory.
This statement is false for a discrete Hopfield network. In the standard discrete Hopfield network, the nodes (or neurons) typically operate using discrete states. These states are usually binary (e.g., 0 or 1) or bipolar (e.g., -1 or +1). A continuous, graded output is a characteristic of the *continuous* Hopfield network, not the discrete version.
This statement is false. A fundamental principle of Hopfield networks (both discrete and continuous) is their energy function. The network dynamics are designed such that the network state evolves towards a stable state, which corresponds to a local minimum of the energy function. Therefore, the energy generally decreases over time, it does not increase.
Based on the analysis of each statement:
Thus, the true statements regarding the discrete Hopfield network are (i) and (ii).
In a multilayer perceptron, an error signal originates at an output neuron of the network and propagates ___________ through the network.
In a feed forward neural network with the following specifications:
Input layer has 4 neurons, hidden layer has 3 neurons and output layer has 2 neurons using the sigmoid activation function for given input values [0.5, 0.8, 0.2, 0.6] as well as the initial weights for the connections.
W1: [0.1, 0.3, 0.5, 0.2]
W2: [0.2, 0.4, 0.6, 0.2] Input layer to hidden layer weights
W3: [0.3, 0.5, 0.7, 0.2]
W4: [0.4, 0.1, 0.3]
W5: [0.5, 0.2, 0.4] Hidden layer to output layer weights
What is the output of the output layer when the given input values are passed through neural network? Round the answer to two decimal places :
Match the following learning modes w.r.t. characteristics of available information for learning :
| a. | Supervised | i. | Instructive information on desired responses, explicitly specified by a teacher. |
| b. | Recording | ii. | A priori design information for memory storing |
| c. | Reinforcement | iii. | Partial information about desired responses, or only “right” or “wrong” evaluative information |
| d. | Unsupervised | iv. | No information about desired responses |
Codes :
a b c d
Consider the following statements about a perception :
I. Feature detector can be any function of the input parameters.
II. Learning procedure only adjusts the connection weights to the output layer.
Identify the correct statement out of the following :
Consider the two class classification task that consists of the following points :
Class \(C_{1}\) : [-1, -1], [-1, 1], [1, -1]
Class \(C_{2}\) : [1, 1]
The decision boundary between the two classes \(C_{1}\) and \(C_{2}\) using single perceptron is given by :