• We covered the use of a this perceptron to calculate the effect of an AND gate on a pair of digital (logical) inputs. 
• We used this diagram in our explanation for to show that a line can separate the (1,1) input from all the other inputs, and therefore the equation of that line can supply the weights and bias to implement the AND function. 
• For the AND function, the weights were: w(1)11 = w(1)12 = 1 and b(1) = -1.5. 
• We also found that by moving the dividing line down, we could separate the input (0,0) at the origin from the other 3 inputs, to accomplish an OR function, with the weights being w(1)11 = w(1)12 = 1 and b(1) = -0.5
• Using this dividing line argument, we found that the XOR function could not be created by a single-layer perceptron because no dividing line could separate the inputs (1,0) and (0,1) from the other two inputs.

Homework/classwork:

1. Convince yourself that there are 16 different possible logical functions (logic gates) with two logical inputs.  Don't forget the function that outputs 1 for every input, and the one that outputs 0 for every input.
2. Convince yourself that, of the 16 possibilities, only 2 of them cannot be accomplished with single-layer perceptrons, with suitable weights and biases.  One of them is, of course, XOR.  What's the other?
3. We can construct an XOR function using a neural net consisting of hidden layer of 2 neurons. Find the appropriate weights (6 of them) and biases (3 of them) to accomplish the XOR.  Note: there is not unique solution to this.  Given a lot of freedom, try to assign the minimum number of different values to the weights and biases (in other words, try to make as many of them the same as possible).
4. Put you results into the Comments-to-Teacher in the homework slot.

Any of you should feel free to answer questions from other students, including coming up to the board to explain.  Just don't give the answers away directly.

There are 4 excellent(!) Youtube videos on how neural networks work from 3Blue1Brown.  Watch at least the first one by the time we return, and more if possible.  Keep the following questions in mind:

• How did he decide on the number of nodes in his network architecture: namely how many hidden layers and the number of nodes in each hidden layer?
  
• How do you calculate the number of weights given the details (above) of the architecture?
  
  
• Do the following calculation:
  • (Calc 1): if his network is presented with a single image, how many operations (each multiplication is an operation, as is an addition) does it take to calculate the single result (the list of 10 probabilities) from that image.  Ignore the calculation of the sigmoid function, and we're only looking for approximate values (+- 10%).
  • (Calc 2): Having done that, now calculate that for a run 30,000 different images. You've just now calculated one data point measuring how well the network classifies (initially, it classifies very poorly). 
  • (Calc 3): Assume that it takes 100,000 data points (trials) to train the network.  If each operation takes a 10 nanoseconds (a single core processor), how long does does training take?  (Ignore the work of backpropagation, if you know about that.)
• Put the answers into the Comments-to-Teacher

• You'll create and submit a small report webpage for this project.
  
• You can work in pairs.  If you do, the report should contain a short discussion of the partnership and division of labor/insight.  Both partners must submit to the homework server.
  
• You need to create at least one, and preferably more than one smarter and faster algorithm. The smartitude of an algorithm smart-algo will be measured by the ratio of your backtracks(naive-algo) / backtracks(smart-algo) for each algo for each board.
• See link below for my webpage format for your report.
• Upload your webpage to your public_html directory, and put the link into the Comments-to-Teacher.  Include in the comments your discussion of the partnership and division of labor.  Also include whether you want your webpage to be seen by others.
  
• Use (Sudoku-boards-Master.txt)

Here's the link page for results.

Copy the answers you wrote on the particular test that was graded into Python and fix your answers until they work.
1. For the first question, use the 3 lines of data on the test as your test file, and see if you get the answers I gave on the test.
2. For the Dlist, you'll have to write both the insert() and delete() methods, as well as __init__() and tolist().  Then do the following:

def insert_all(the_dlist, the_input_list):
  for element in the_input_list:
    the_dlist.insert(element)
a = Dlist()
insert_all(a,[4,5,2,3,2,7])
print(a.tolist())

and you should get: [2,2,3,4,5,7]

['B', 'E', 'E', 'R', 'T', 'R', 'P']
['E', 'P', 'R', 'R', 'T']

1. The Turing Test in the original: Computing Machinery and Intelligence by Alan Turing -- this is (possibly) the most famous paper about AI.  Read sections 1, 2 and 6; the other sections are optional.
2. AlphaZero: read the first two pages of the research paper announcing AlphaZero -- this is the type of publishable paper you'll find in AI research. BTW, for those of you who have Netflix accounts, there is a documentary there called ... "AlphaGo" about .... AlphaGo.

Hey folks!
My name is Joan Chirinos, and I'm the president of the Stuy Competitive Computing Club.

"What's the Competitive Computing Club???"

Good question! As upperclassmen, we will teach you how to use your rad CS skills to solve the problems typically found in CS competitions. 

We will also use this knowledge in a number of fun activities, including mini-competitions! 

With your immense knowledge, you can even go to PClassic or SJC, larger and more structured CS competitions!

If this sounds interesting to you, please fill out the form below. I will be emailing you shortly about our first meeting!

https://goo.gl/forms/GvWdAK0RryQZDIm82 

Here's the BinTree notebook to complete.  Once completed and tested, download the HTML version of your notebook and submit to the homework server.

Slight change in the notebook above: the tree should have no duplicates.

Here's my answer: BinTree

Dr. Nikolaus Kriegeskorte is a professor of neuroscience and a Principal Investigator and Director of Cognitive Imaging at the Zuckerman Mind Brain Behavior Institute at Columbia University. He uses deep neural network models, a form of AI, to mirror the brain’s visual system. Over the years, Dr. Kriegeskorte and his team have been improving their systems to mimic biological processes with greater accuracy. Moreover, these AI models not only simulate vision system but also other types of intelligence, such as language, reasoning, and motor control.

Website:

1. Go to the main class link page (http://bert.stuy.edu/pbrooks)
2. Click on the Homework/Grade server link
3. Choose the Profile menu item
4. Choose your period and name
5. Log in with your OSIS number as your password
6. Fill out your email address and preferred first name only, DO NOT CHANGE YOUR PASSWORD YET.
7. Save: (Change Profile button)...
8. (optional, but recommended): now you can change your password for this Homework server, if you want.
9. REMEMBER this password!