3

.pdf

School

Nanyang Technological University *

*We aren’t endorsed by this school

Course

1701X

Subject

Computer Science

Date

Nov 24, 2024

Type

pdf

Pages

1

Uploaded by DukeLemur2884

Report
Classification Bank Notes Workshop Overview In this workshop, we train a neural network to classify a given banknote as either Genuine or Counterfeit. Data The dataset provided is called ‘banknotes.csv’ and consists of physical dimensions of genuine and counterfeit Swiss banknotes. There are 200 samples in the dataset; 100 are samples of Genuine banknotes and 100 are samples of Counterfeit banknotes. Source of dataset - http:// www.statistics4u.com/fundstat_eng/data_fluriedw.html Tasks 1. Use Pandas to read in ‘banknotes.csv’. 2. Use the following columns as features : Length Left Right Bottom Top Diagonal 3. Use the column ‘Genuine’ as your label , where a value of 0 denotes a Counterfeit sample, and a value of 1 denotes a Genuine sample. For example, with reference to the snapshot below, the sample BN100 denotes a Genuine banknote, while BN101 denotes a Counterfeit banknote. 4. Out of the 200 samples, use only 90 Genuine and 90 Counterfeit samples for training. In other words, your training features consists of 180 rows of data (half are Genuine, the other half Counterfeit). Naturally, your training labels would then come from the selected 180 rows of data. The remaining 20 samples (10 Genuine; 10 Counterfeit) is for testing the accuracy of your neural network. 5. Construct your neural network model and train it with 180 training samples. Then evaluate its accuracy with the remaining 20 test samples.
Discover more documents: Sign up today!
Unlock a world of knowledge! Explore tailored content for a richer learning experience. Here's what you'll get:
  • Access to all documents
  • Unlimited textbook solutions
  • 24/7 expert homework help