Signal Processing - November 2017 - 119

where, {b 1, f, b K } are a set of basis vectors that are learned
from unlabeled data and b i ! R d . For input data x iu, the corj
responding sparse representation is a i = {a 1i , f, a Ki }, with a i
corresponding to the basis vector bj . The transfer of learning
occurs when the same set of basis vectors {b 1, f, b K } is used
as a basis to represent labeled target data. Some of the prominent machine-learning and computer vision techniques that
incorporate self-taught learning are [15]-[18].

Sample selection bias
The concept of sample selection bias was introduced in economics as a Nobel prize-winning work by James Heckman in 1979
[19]. When a distribution of sampled data does not reflect the
true distribution of the data set it is sampled from, it is a case
of sample selection bias. For example, a financial bank intends
to model the profile of a loan defaulter to deny such defaulters a loan from the bank. It therefore builds a model based
on the loan defaulters it has in its records. However, this is a
small subset and, therefore, does not truthfully model the general public the bank wants to profile but does not have access
to. Therefore, the defaulter profile generated by the bank is offset by what is called the sample selection bias. In this learning
scenario, a data set X = {x i, y i}in= 1 is made available. This data
set is used to estimate the joint distribution Pt (X, Y ), which is an
approximation for the true joint distribution P (X, Y ) . However,
Pt (X, Y ) ! P (X, Y ), where Pt (X, Y ) is the estimated distribution
and P (X, Y ) is the true distribution. This could be because there
are very few data samples, which could lead to a poor estimation
of the prior distribution, Pt (X ) ! P (X ) . Other cases when the
training data does not represent the target (test) data and introduces a bias in the class prior (Pt (Y ) ! P (Y )) eventually lead to
incorrect estimation of the conditional (Pt (Y | X ) ! P (Y | X )) . To
correct this discrepancy, knowledge transfer is implemented by
weighting the training data samples to reflect the test distribution
[20]. When both the marginal (Pt (X ) ! P (X )) and the conditionals are different (Pt (Y | X ) ! P (Y | X )), the problem is referred to
as sample selection bias [21]-[23].

LML
The concept of lifelong learning was discussed in the seminal
work by Thrun [24]. The concept of transfer in lifelong learning can be formulated as follows. A machine-learning model
trained for K tasks {T1, T2, f, TK } is updated by learning
task TK +1 with data D K +1 . The work discussed if learning
the K + 1th task was easier than learning the first task. The key
characteristics of lifelong learning are 1) a continuous learning
process, 2) knowledge accumulation, and 3) use of past knowledge to assist in future learning [25]. LML differs from MTL
because it retains knowledge about previous tasks and applies
that knowledge to learn new tasks. It also differs from standard
domain adaptation, which transfers knowledge to learn only one
task (target).

One-shot learning and zero-shot learning
These can be viewed as extreme cases of transfer learning [11].
Both these forms of transfer seek to learn data categories from

minimal data. The key motivation is the ability to transfer knowledge from previously learned categories to recognize new categories. In one-shot learning, the model is trained to recognize a new
category of data based on just one labeled example [26]. It relies
on the ability of the model to learn representations that cleanly
separate the underlying categories. On the other hand, zeroshot learning is the ability to recognize new -categories without
having seen any example of them. Zero-data learning [27] and
zero-shot learning [28], [29] are examples where the model has
learned to transfer knowledge from training data not completely
related to the categories of interest. For example, a model that has
been trained to recognize breeds of dogs can be provided with
a description of the categories {fox, wolf, hyena, wild dog}.
Without having ever seen an image of any of these categories, the
zero-data learning model can be trained to associate the textual
description to learn and recognize the new category.

Domain adaptation
In domain adaptation, the source domain D S and the target
domain D T are not the same, and the goal is to solve a common task T = {Y, f ($)}. For example, in an image-recognition
task, the source domain could contain labeled images of objects
against a white background, and the target domain could consist of unlabeled images of objects against a noisy and cluttered background. Both the domains inherently have the same
set of image categories. The difference between the domains is
modeled as the variation in their joint probability distributions
PS (X, Y ) ! PT (X, Y ) [10]. Standard domain adaptation assumes
that there is plenty of labeled data in the source domain and there
is no labeled data (or few samples) in the target domain. Since
there are no labeled samples (or very few) of target data, it is difficult to get a good estimate of PtT (X, Y ). The key task of domain
adaptation lies in approximating Pt T (X, Y ) using the source data
distribution estimation Pt S (X, Y ). This is possible because the
two domains are assumed to be correlated. This correlation is
often modeled as covariate shift, where PS (X ) ! PT (X ) and
PS (Y | X ) . PT (Y | X ).
Domain adaptation algorithms are evaluated on the basis of
minimizing the expected error of prediction on the target data set.
A classification model for the target is usually trained using the
source data, which has labels along with the target data without
labels (or very few labels). Domain adaptation algorithms can be
classified largely into two groups based on the availability of labels
for the target data. In unsupervised domain adaptation, there are
no labels for the target data. Only the source data has labels [30]-
[33]. A classifier trained with the labeled source data are adapted
to the unlabeled target data. In supervised domain adaptation, a
few labeled samples are present in the target domain for all of
the categories. However, these are few in number, and a target
classifier trained with only these data points could overfit. This
paradigm is also referred to as semisupervised domain adaptation
because there are labels for only a few target samples in each category and the unlabeled target data is also used in a transductive
setting to estimate the labels [34]-[36]. The source data, which
have much more labeled data, are used along with the target data
to estimate the optimal target classifier and prevent overfitting.

IEEE SIGNAL PROCESSING MAGAZINE

|

November 2017

|

119



Table of Contents for the Digital Edition of Signal Processing - November 2017

Signal Processing - November 2017 - Cover1
Signal Processing - November 2017 - Cover2
Signal Processing - November 2017 - 1
Signal Processing - November 2017 - 2
Signal Processing - November 2017 - 3
Signal Processing - November 2017 - 4
Signal Processing - November 2017 - 5
Signal Processing - November 2017 - 6
Signal Processing - November 2017 - 7
Signal Processing - November 2017 - 8
Signal Processing - November 2017 - 9
Signal Processing - November 2017 - 10
Signal Processing - November 2017 - 11
Signal Processing - November 2017 - 12
Signal Processing - November 2017 - 13
Signal Processing - November 2017 - 14
Signal Processing - November 2017 - 15
Signal Processing - November 2017 - 16
Signal Processing - November 2017 - 17
Signal Processing - November 2017 - 18
Signal Processing - November 2017 - 19
Signal Processing - November 2017 - 20
Signal Processing - November 2017 - 21
Signal Processing - November 2017 - 22
Signal Processing - November 2017 - 23
Signal Processing - November 2017 - 24
Signal Processing - November 2017 - 25
Signal Processing - November 2017 - 26
Signal Processing - November 2017 - 27
Signal Processing - November 2017 - 28
Signal Processing - November 2017 - 29
Signal Processing - November 2017 - 30
Signal Processing - November 2017 - 31
Signal Processing - November 2017 - 32
Signal Processing - November 2017 - 33
Signal Processing - November 2017 - 34
Signal Processing - November 2017 - 35
Signal Processing - November 2017 - 36
Signal Processing - November 2017 - 37
Signal Processing - November 2017 - 38
Signal Processing - November 2017 - 39
Signal Processing - November 2017 - 40
Signal Processing - November 2017 - 41
Signal Processing - November 2017 - 42
Signal Processing - November 2017 - 43
Signal Processing - November 2017 - 44
Signal Processing - November 2017 - 45
Signal Processing - November 2017 - 46
Signal Processing - November 2017 - 47
Signal Processing - November 2017 - 48
Signal Processing - November 2017 - 49
Signal Processing - November 2017 - 50
Signal Processing - November 2017 - 51
Signal Processing - November 2017 - 52
Signal Processing - November 2017 - 53
Signal Processing - November 2017 - 54
Signal Processing - November 2017 - 55
Signal Processing - November 2017 - 56
Signal Processing - November 2017 - 57
Signal Processing - November 2017 - 58
Signal Processing - November 2017 - 59
Signal Processing - November 2017 - 60
Signal Processing - November 2017 - 61
Signal Processing - November 2017 - 62
Signal Processing - November 2017 - 63
Signal Processing - November 2017 - 64
Signal Processing - November 2017 - 65
Signal Processing - November 2017 - 66
Signal Processing - November 2017 - 67
Signal Processing - November 2017 - 68
Signal Processing - November 2017 - 69
Signal Processing - November 2017 - 70
Signal Processing - November 2017 - 71
Signal Processing - November 2017 - 72
Signal Processing - November 2017 - 73
Signal Processing - November 2017 - 74
Signal Processing - November 2017 - 75
Signal Processing - November 2017 - 76
Signal Processing - November 2017 - 77
Signal Processing - November 2017 - 78
Signal Processing - November 2017 - 79
Signal Processing - November 2017 - 80
Signal Processing - November 2017 - 81
Signal Processing - November 2017 - 82
Signal Processing - November 2017 - 83
Signal Processing - November 2017 - 84
Signal Processing - November 2017 - 85
Signal Processing - November 2017 - 86
Signal Processing - November 2017 - 87
Signal Processing - November 2017 - 88
Signal Processing - November 2017 - 89
Signal Processing - November 2017 - 90
Signal Processing - November 2017 - 91
Signal Processing - November 2017 - 92
Signal Processing - November 2017 - 93
Signal Processing - November 2017 - 94
Signal Processing - November 2017 - 95
Signal Processing - November 2017 - 96
Signal Processing - November 2017 - 97
Signal Processing - November 2017 - 98
Signal Processing - November 2017 - 99
Signal Processing - November 2017 - 100
Signal Processing - November 2017 - 101
Signal Processing - November 2017 - 102
Signal Processing - November 2017 - 103
Signal Processing - November 2017 - 104
Signal Processing - November 2017 - 105
Signal Processing - November 2017 - 106
Signal Processing - November 2017 - 107
Signal Processing - November 2017 - 108
Signal Processing - November 2017 - 109
Signal Processing - November 2017 - 110
Signal Processing - November 2017 - 111
Signal Processing - November 2017 - 112
Signal Processing - November 2017 - 113
Signal Processing - November 2017 - 114
Signal Processing - November 2017 - 115
Signal Processing - November 2017 - 116
Signal Processing - November 2017 - 117
Signal Processing - November 2017 - 118
Signal Processing - November 2017 - 119
Signal Processing - November 2017 - 120
Signal Processing - November 2017 - 121
Signal Processing - November 2017 - 122
Signal Processing - November 2017 - 123
Signal Processing - November 2017 - 124
Signal Processing - November 2017 - 125
Signal Processing - November 2017 - 126
Signal Processing - November 2017 - 127
Signal Processing - November 2017 - 128
Signal Processing - November 2017 - 129
Signal Processing - November 2017 - 130
Signal Processing - November 2017 - 131
Signal Processing - November 2017 - 132
Signal Processing - November 2017 - 133
Signal Processing - November 2017 - 134
Signal Processing - November 2017 - 135
Signal Processing - November 2017 - 136
Signal Processing - November 2017 - 137
Signal Processing - November 2017 - 138
Signal Processing - November 2017 - 139
Signal Processing - November 2017 - 140
Signal Processing - November 2017 - 141
Signal Processing - November 2017 - 142
Signal Processing - November 2017 - 143
Signal Processing - November 2017 - 144
Signal Processing - November 2017 - 145
Signal Processing - November 2017 - 146
Signal Processing - November 2017 - 147
Signal Processing - November 2017 - 148
Signal Processing - November 2017 - 149
Signal Processing - November 2017 - 150
Signal Processing - November 2017 - 151
Signal Processing - November 2017 - 152
Signal Processing - November 2017 - 153
Signal Processing - November 2017 - 154
Signal Processing - November 2017 - 155
Signal Processing - November 2017 - 156
Signal Processing - November 2017 - 157
Signal Processing - November 2017 - 158
Signal Processing - November 2017 - 159
Signal Processing - November 2017 - 160
Signal Processing - November 2017 - 161
Signal Processing - November 2017 - 162
Signal Processing - November 2017 - 163
Signal Processing - November 2017 - 164
Signal Processing - November 2017 - 165
Signal Processing - November 2017 - 166
Signal Processing - November 2017 - 167
Signal Processing - November 2017 - 168
Signal Processing - November 2017 - 169
Signal Processing - November 2017 - 170
Signal Processing - November 2017 - 171
Signal Processing - November 2017 - 172
Signal Processing - November 2017 - 173
Signal Processing - November 2017 - 174
Signal Processing - November 2017 - 175
Signal Processing - November 2017 - 176
Signal Processing - November 2017 - Cover3
Signal Processing - November 2017 - Cover4
https://www.nxtbook.com/nxtbooks/ieee/signalprocessing_201809
https://www.nxtbook.com/nxtbooks/ieee/signalprocessing_201807
https://www.nxtbook.com/nxtbooks/ieee/signalprocessing_201805
https://www.nxtbook.com/nxtbooks/ieee/signalprocessing_201803
https://www.nxtbook.com/nxtbooks/ieee/signalprocessing_201801
https://www.nxtbook.com/nxtbooks/ieee/signalprocessing_1117
https://www.nxtbook.com/nxtbooks/ieee/signalprocessing_0917
https://www.nxtbook.com/nxtbooks/ieee/signalprocessing_0717
https://www.nxtbook.com/nxtbooks/ieee/signalprocessing_0517
https://www.nxtbook.com/nxtbooks/ieee/signalprocessing_0317
https://www.nxtbook.com/nxtbooks/ieee/signalprocessing_0117
https://www.nxtbook.com/nxtbooks/ieee/signalprocessing_1116
https://www.nxtbook.com/nxtbooks/ieee/signalprocessing_0916
https://www.nxtbook.com/nxtbooks/ieee/signalprocessing_0716
https://www.nxtbook.com/nxtbooks/ieee/signalprocessing_0516
https://www.nxtbook.com/nxtbooks/ieee/signalprocessing_0316
https://www.nxtbook.com/nxtbooks/ieee/signalprocessing_0116
https://www.nxtbook.com/nxtbooks/ieee/signalprocessing_1115
https://www.nxtbook.com/nxtbooks/ieee/signalprocessing_0915
https://www.nxtbook.com/nxtbooks/ieee/signalprocessing_0715
https://www.nxtbook.com/nxtbooks/ieee/signalprocessing_0515
https://www.nxtbook.com/nxtbooks/ieee/signalprocessing_0315
https://www.nxtbook.com/nxtbooks/ieee/signalprocessing_0115
https://www.nxtbook.com/nxtbooks/ieee/signalprocessing_1114
https://www.nxtbook.com/nxtbooks/ieee/signalprocessing_0914
https://www.nxtbook.com/nxtbooks/ieee/signalprocessing_0714
https://www.nxtbook.com/nxtbooks/ieee/signalprocessing_0514
https://www.nxtbook.com/nxtbooks/ieee/signalprocessing_0314
https://www.nxtbook.com/nxtbooks/ieee/signalprocessing_0114
https://www.nxtbook.com/nxtbooks/ieee/signalprocessing_1113
https://www.nxtbook.com/nxtbooks/ieee/signalprocessing_0913
https://www.nxtbook.com/nxtbooks/ieee/signalprocessing_0713
https://www.nxtbook.com/nxtbooks/ieee/signalprocessing_0513
https://www.nxtbook.com/nxtbooks/ieee/signalprocessing_0313
https://www.nxtbook.com/nxtbooks/ieee/signalprocessing_0113
https://www.nxtbook.com/nxtbooks/ieee/signalprocessing_1112
https://www.nxtbook.com/nxtbooks/ieee/signalprocessing_0912
https://www.nxtbook.com/nxtbooks/ieee/signalprocessing_0712
https://www.nxtbook.com/nxtbooks/ieee/signalprocessing_0512
https://www.nxtbook.com/nxtbooks/ieee/signalprocessing_0312
https://www.nxtbook.com/nxtbooks/ieee/signalprocessing_0112
https://www.nxtbook.com/nxtbooks/ieee/signalprocessing_1111
https://www.nxtbook.com/nxtbooks/ieee/signalprocessing_0911
https://www.nxtbook.com/nxtbooks/ieee/signalprocessing_0711
https://www.nxtbook.com/nxtbooks/ieee/signalprocessing_0511
https://www.nxtbook.com/nxtbooks/ieee/signalprocessing_0311
https://www.nxtbook.com/nxtbooks/ieee/signalprocessing_0111
https://www.nxtbook.com/nxtbooks/ieee/signalprocessing_1110
https://www.nxtbook.com/nxtbooks/ieee/signalprocessing_0910
https://www.nxtbook.com/nxtbooks/ieee/signalprocessing_0710
https://www.nxtbook.com/nxtbooks/ieee/signalprocessing_0510
https://www.nxtbook.com/nxtbooks/ieee/signalprocessing_0310
https://www.nxtbook.com/nxtbooks/ieee/signalprocessing_0110
https://www.nxtbook.com/nxtbooks/ieee/signalprocessing_1109
https://www.nxtbook.com/nxtbooks/ieee/signalprocessing_0909
https://www.nxtbook.com/nxtbooks/ieee/signalprocessing_0709
https://www.nxtbook.com/nxtbooks/ieee/signalprocessing_0509
https://www.nxtbook.com/nxtbooks/ieee/signalprocessing_0309
https://www.nxtbook.com/nxtbooks/ieee/signalprocessing_0109
https://www.nxtbook.com/nxtbooks/ieee/signalprocessing_1108
https://www.nxtbook.com/nxtbooks/ieee/signalprocessing_0908
https://www.nxtbook.com/nxtbooks/ieee/signalprocessing_0708
https://www.nxtbook.com/nxtbooks/ieee/signalprocessing_0508
https://www.nxtbook.com/nxtbooks/ieee/signalprocessing_0308
https://www.nxtbook.com/nxtbooks/ieee/signalprocessing_0108
https://www.nxtbookmedia.com