Machine Learning & Data Mining
CS/CNS/EE 155
Lecture 9:
Condi5onal Random Fields
1
Announcements
Homework 5 released
Skeleton code available on Moodle
Due in 2 weeks (2/16)
Kaggle compe55on closes 2/9
SHORT r
Machine Learning & Data Mining
Homework 2
Caltech CS/CNS/EE 155
January 12th , 2016
This set is due 2pm, January 19th , via Moodle. You are free to collaborate on all of the problems, subject
to the collaboration policy stated in the syllabus. Please incl
Physics 125b Problem Set 6 Due March 6, 2014
Problem 1
(10 points)
Suppose the scattering amplitude for a certain reaction is given by
1
k
3
f () =
+ 3e2ik sin(2k 3 ) cos ,
k k0 k ik
where , k0 andp
are real constants characteristic of the potential whic
Physics 125b Problem Set 3 Due January 30, 2014
Problem 1
(20 points)
The charge of the nucleus of an atom Z(t) changes from Z0 to (Z0 1) as time goes from t = 0 to
t = in the following fashion,
Z(t) = (Z0 1) + et/ .
Treating the second term above as a pe
Ph 125B Problem Set 4
Wentao Fan
1).
(a)
1
2
1=
3
2
1
2
For the subspace where j =
3
2,
we have | 23 , 32 i = | 12 ; 1i by the Condon and Shortley phase convention, and this gives the first
non-zero Clebsch-Gordan coefficient h 12 ; 1| 32 , 32 i = 1. From
Ph 125B Problem Set 1
Wentao Fan
1).
We have shown in the lectures that when we act the raising operator L+ Lx + iLy on an energy eigenstate |n, l, mi, we get
p
L+ |n, l, mi = ~ (l m)(l + m + 1)|n, l, m + 1i. To show that states with the same set of quant
Machine Learning & Data Mining
CS/CNS/EE 155
Lecture 10:
Condi6onal Random Fields Revisited,
Overview of General Structured Predic6on
1
Today
Nave Bayes vs Logis6c Regression
Detailed Comparison
Generalizes Conceptually to HMMs vs CRFs
Condi
Machine Learning & Data Mining
Miniproject 2: Shakespearebot 5000
1
Caltech CS/CNS/EE 155
February 25th , 2016
Overview
In this project, we will attempt to generate Shakespearean sonnets by training a HMM on the entire
corpus of Shakespeares sonnets.
Th
Physics 125b
Problem Set 3, Due Thursday. Jan 28, 2016
Note: The rst problem relates to work from last term.
Problem 1
Show that the quantization condition for l = 0 bound states in a spherical
well of depth V0 and radius r0 is
k
= tan(kr0 )
(1)
where k i
Physics 125b
Problem Set 1, Due Wed. Jan 13, 2016
Problem 1
Derive the Clebsh-Gordon coecients for,
1
2
1=
3
2
1
2
Problem 2
Derive the Clebsh-Gordon coecients for, 1 1 = 2 1 0
Problem 3
Consider a hydrogen atom with an additional angular momentum depende
Machine Learning & Data Mining
Homework 5
Caltech CS/CNS/EE 155
February 2th
This set is due 2:30pm, February 16th , via Moodle. Note that you get two weeks for this set. You are free
to collaborate on all of the problems, subject to the collaboration pol
Machine Learning & Data Mining
Homework 4
Caltech CS/CNS/EE 155
January 26th
This set is due 2:30pm, February 2th , via Moodle. You are free to collaborate on all of the problems,
subject to the collaboration policy stated in the syllabus.
1
Sequence Pred
Machine Learning & Data Mining
Homework 6
Caltech CS/CNS/EE 155
February 16th
This set is due 2:30pm, March 1st , via Moodle. Note that you get two weeks for this set. You are free to
collaborate on all of the problems, subject to the collaboration policy
Machine Learning & Data Mining
Homework 4
Caltech CS/CNS/EE 155
January 26th
This set is due 2:30pm, February 2th , via Moodle. You are free to collaborate on all of the problems,
subject to the collaboration policy stated in the syllabus.
1
Sequence Pred
Machine Learning & Data Mining
CS/CNS/EE 155
Lecture 11:
Recent Applica6ons
1
Kaggle Miniproject Closed
2
1
dge Detection Using Structured Forests
Today
and C. Lawrence Zitnick
Piotr Dollar
Microsoft Research
cfw_pdollar,[email protected]
Physics 125b Problem Set 5 due February 27, 2014
Problem 1
(5 points)
A typical weak cross section at LHC energies is a picobarn. Suppose the cross section for scattering
is 1 picobarn = 1012 barn = 1040 meters2 . The beam of incident particles is moving