Question$-1(12$ points$)$

According to Metcalf Theory, the value of a network depends on the number of connections within this network. This is what is called $$Network Effects$.$

Assume that we have two networks that are competing with each other in the online dating arena: One of them has $1,000$ users, and the other has $2,000$ users.

Determine the value of each dating service independently based on its number of connections, and the value of the network if both dating services are merged.

$.$

$$

Question$-2(8$ points$)$

Which four of the following data features are considered characteristics of Big Data?

Characteristics Big Data Not big Data

Speed of the Data

Language of the Data

Reliability of the Data

Differences in the Types of the Data

Normalization of the Data

Size of the Data

Consistent Structure of the Data

Distribution of the Data

$$

Question$-3(16$ points$)$

Link the following solutions to the Neural Network problems identified below:

$$ Principal Component Analysis

$$ Using More Training Data

$$ Increase the Model Complexity

$$ Implement Gradient Clipping

$$ Neuromorphic Computing

$$ Use of GPUs and ASICs

$$ Conversion to Local Optimized Solution

$$ Use of Multiple Sources of Data

$$ Boltzmann Machine

Neural Network Problem Solution

Underfitting

Overfitting

Neural Network Training is Resource Hungry

Poor Data Quality

Current Hardware was not Designed for Neural Networks

Diminishing Gradients

Avoid Initializing all Weights to the Same Value

Too Many Dimensions

$$

Question$-4(12$ points$)$:

Identify the type of network effects associated with each of these companies.

Alternatives are:

$$ One$-$sided Network Effects

$$ Two$-$sided Network Effects

$$ Three$-$sided Network Effects

$$ No Network Effects

Company $/$ Line of Business Network Effects

Disney Plus

Door Dash

Dell Computers $($hardware sales$)$

Instagram

WeChat Pay

YouTube

$$

Question$-5(12$ points$)$

Match the applications with the following architectures$/$technologies:

$$ Client$-$Server

$$ Three$-$tier Architecture

$$ Super Computing

$$ Virtualization

$$ Open Source

$$ Quantum Computing

$$ LAMP

$$ Neuromorphic Computing

$$ RDBMS

$$ Grid$/$Cluster Computing

Please provide a justification for your choice.

Application Architecture

AI Data Training for Self$-$Driving

AI Process Execution for Self$-$Driving

Bitcoin Mining

The Internet

Seismic Event Forecasting

Genetic Analysis