Question

I have some questions related to machine learning course : how can I use k-NN classification...

I have some questions related to machine learning course : how can I use k-NN classification with Olivetti Faces dataset ?

0 0
Add a comment Improve this question Transcribed image text
Answer #1

Answer:-

In machine learning, we represent each data instances in a feature space. Refer: Mark Whitfield Iris flower dataset for example. So, each data point co-ordinates corresponds to the feature. With this representation, you can apply KNN to any dataset.

face recognition was performed using the face images in the Olivetti data set. The steps for face recognition are as follows:

  • Principal components of face images were obtained by PCA.
  • Adequate number of principal components determined
  • According to three different classification models, accuracy score obtained.
  • According to three different classification models, cross-validation accuracy score were obtained.
  • Parameter optimization of the best model has been made.

We can classify the face recognition researches carried out with 2D face recognition approach in three categories; analytical (feature-based, local), global (appearance) and hybrid methods. While analytical approaches want to be recognized by comparing the properties of the facial components, global approaches try to achieve a recognition with data derived from all the face. Hybrid approaches, together with local and global approaches, try to obtain data that expresses the face more accurately.

Face recognition performend in this kernel can evaluated under global face recognition approaches.

In analytical approaches, the distance of the determined feature points and the angles between them, the shape of the facial features or the variables containing the regional features are obtained from the face image and the variables are used in face recognition. Analytical methods examine the face images in two different ways according to the pattern and geometrical properties. In these methods, the face image is represented by smaller size data, so the big size problem that increases the operations in face recognition is solved.

  • There are ten different image of each of 40 distinct people
  • There are 400 face images in the dataset
  • Face images were taken at different times, variying ligthing, facial express and facial detail
  • All face images have black background
  • The images are gray level
  • Size of each image is 64x64
  • Image pixel values were scaled to [0, 1] interval
  • Names of 40 people were encoded to an integer from 0 to 39
Add a comment
Know the answer?
Add Answer to:
I have some questions related to machine learning course : how can I use k-NN classification...
Your Answer:

Post as a guest

Your Name:

What's your source?

Earn Coins

Coins can be redeemed for fabulous gifts.

Not the answer you're looking for? Ask your own homework help question. Our experts will answer your question WITHIN MINUTES for Free.
Similar Homework Help Questions
ADVERTISEMENT
Free Homework Help App
Download From Google Play
Scan Your Homework
to Get Instant Free Answers
Need Online Homework Help?
Ask a Question
Get Answers For Free
Most questions answered within 3 hours.
ADVERTISEMENT
ADVERTISEMENT