Q1 Gradient descent is a very helpful algorithm. But it is not guaranteed to converge to global
minima always. Give an example of a continuous function and initial point for which gradient
Study Session 11 builds on the fundamentals of fixed-income portfolio management
to address international and emerging market strategies and the use of derivatives to
manage interest rate and credit
Effective risk management identifies, assesses, and controls numerous sources of
risk, both financial and nonmarket related, in an effort to achieve the highest possible level of reward for the risks incurred. With the in
The Global Investment Performance Standards (GIPS) contain ethical and profes-
sional standards for presenting investment performance to prospective clients. These
guidelines provide for standardiz
Asset Allocation and Related
Decisions in Portfolio
When the strategic asset allocation includes exposure to global markets, non-
domestic currencies create an additional source of portfolio volatility. The investment
Equity Portfolio Management
Because equity securities represent a significant portion of many investment port-
folios, equity portfolio management is often an important component of overall
investment success. This study session focuses
Performance evaluation addresses three questions that are essential in evaluating the
results of the portfolio management process:
What was the portfolios performance?
Why did the portfolio produce the observed per
The elements of the investment management process, such as the consideration of
risk, return, and investment constraints, are the same for a fixed-income portfolio as
for any other type of portfolio.
Alternative Investments for
Alternative investments comprise groups of investments with risk and return char-
acteristics that differ markedly from those of traditional stock and bond investments.
Common features of
Applications of Derivatives
This study session addresses risk management strategies using forwards and futures,
option strategies, floors and caps, and swaps. These derivatives can be used for a
variety of risk management
Because the investment process is not complete until securities are bought or sold,
the quality of trade execution is an important determinant of investment results. The
methods by which managers and
Introduction to Machine Learning - CS419
Instructor: Prof. Ganesh Ramakrishnan
Lecture 2 - Supervised vs. Unsupervised Learning
and Method of Least Squares
Supervised vs Unsupervised
Task: Suppose you had a basket and it is fulled with some fresh
No marks. This quiz is only to help us understand your background
and help you understand desirable bakcground knowledge. The TAs
will help you pick up these basics if you need BUT you will need to
work as well.
1. Suppose that you have two differe
Friday 11th March, 2016
Problem 1. Design a multilayer perceptron which will learn to recognize various forms of
the the letters C,L,T placed on a 3x3 grid through backpropagation algorithm.
1. Design a one layer network indicating what should
Detecting spam mails
One of the fundamental tasks of machine learning is to detect spam e-mails.
You are given some words and a label of +1 if it is spam or -1 if it is not. Here 1
indicates the presence of word and 0 the absence of word. Assum
Basis function expansion & Kernel: Part 1
We saw the that for p [0, ), under certain conditions on K, the
following can be equivalent representations
wj j (x)
i K(x, xi )
For what kind of regularizers, loss functions and p [0
Introduction to Machine Learning - CS403/725
Instructor: Prof. Ganesh Ramakrishnan
Lecture 1 : Introduction and Motivation
Introduction: What is Machine Learning?
Machine learning is a sub-eld of computer science that
evolved from the study of pattern rec
Friday 29th January, 2016; 23:17
Problem 1. Which of the following sets are convex?
1. A slab, i.e., a set of the form cfw_x Rn | aT x .
2. A rectangle, i.e., a set of the form cfw_x Rn |i xi i , i = 1, . . . , n.
3. A wedge, i.e., cfw_x Rn |aT
Closed notes, 30 Marks, 2 hours
Tuesday 1st March, 2016
Problem 1. Support Vector Regression:
1. If all training data points lie strictly inside the -band of the SVR solution, what
would the regression line be? You can build upon basic