A car company is developing a machine learning solution to detect whether a car is present in an image. The image dataset consists of one million images. Each image in the dataset is 200 pixels in…

Question

Options

• AUse a deep convolutional neural network (CNN) classifier with the images as input. Include a
• BUse a deep convolutional neural network (CNN) classifier with the images as input. Include a
• CUse a deep multilayer perceptron (MLP) classifier with the images as input. Include a linear
• DUse a deep multilayer perceptron (MLP) classifier with the images as input. Include a softmax

• A
  3% (1)
• B
  84% (31)
• C
  8% (3)
• D
  5% (2)

Explanation

Convolutional Neural Networks (CNNs) are the gold-standard architecture for image classification tasks. CNNs use convolutional layers to automatically learn spatial hierarchies of features (edges, textures, shapes, objects) from raw pixel data, making them far superior to MLPs for visual tasks. A deep CNN (B) with a softmax output layer is the correct architecture: the convolutional and pooling layers extract discriminative visual features from the 200×200 pixel images, and the softmax layer produces a probability distribution over the two classes (car / no car), enabling confident binary classification. Option A may differ in output activation (e.g., sigmoid vs. softmax). Options C and D use Multilayer Perceptrons, which treat each pixel as an independent feature, losing all spatial structure - MLPs perform significantly worse on image data than CNNs. Option C uses a linear output, which is inappropriate for classification.

No community discussion yet for this question.