Coms 6998-3: Sub-linear Algorithms in Learn- Ing and Testing

Coms 6998-3: Sub-linear Algorithms in Learn- Ing and Testing Spring 2014 2 Learning Boolean Functions 2.1 Setup

We describe the framework for uniform-distribution learning of a class C of Boolean functions f : {0, 1} → {0, 1} using membership queries (MQ). The class C should be thought of as “known” to the learning algorithm; there is also an unknown target function f ∈ C. The learning algorithm has MQ access to f . Recall that given f, g : {0, 1} → {0, 1}, the distance dist(f, g) is the probability over...

Notes for COMS 6998-4 Fall 2017

v 3 COMS E 6998 - 002 : Probabilistic Modeling for Discrete Data Lecture 3 : Word Embeddings III Instructor

If I missed any of your comments from class below or misattributed an existing comment, it was not intentional, and I'll fix it. The following scribed notes are provided in a stream of consciousness style. If anything is missing or confusing, please let me know and I will update the notes. 0 Logistics • If you are using GitHub, please switch to BitBucket. • Please update your project log every ...

COMS 6998 : Advanced Complexity Spring 2017 Lecture 2 : January 26 , 2017

1. The Andre'ev (1987) bound: explicit function with size f = ˜ Ω(n 5 2) 2. The relationship between formula size and circuit depth, and how to accomplish " depth reduction " for a general circuit 3. Monotone formulas for monotone functions (example: monotone formula for the majority function MAJ)

COMS 6998 - 4 Fall 2017 Presenter : Wenxi Chen

This lecture is delivered in more philosophical sense rather than technically. In the past, we have learned a series of algorithms which can dig deeper in a training dataset and generate a model based on such a dataset. However, without knowing of what knowledge is contained in the model, it is generally hard for human beings to trust the trained model and apply it in reality. Sometimes, biased...