In logic, statistical inference, and supervised learning,transduction or transductive inference is reasoning fromobserved, specific (training) cases to specific (test) cases. In contrast,induction is reasoning from observed training casesto general rules, which are then applied to the test cases. The distinction ismost interesting in cases where the predictions of the transductive model arenot achievable by any inductive model. Note that this is caused by transductiveinference on different test sets producing mutually inconsistent predictions.
Transduction was introduced in a computer science context by Vladimir Vapnik in the 1990s, motivated byhis view that transduction is preferable to induction since, according to him, induction requiressolving a more general problem (inferring a function) before solving a morespecific problem (computing outputs for new cases): "When solving a problem ofinterest, do not solve a more general problem as an intermediate step. Try toget the answer that you really need but not a more general one."[1] .
An example of learning which is not inductive would be in the case of binaryclassification, where the inputs tend to cluster in two groups. A large set oftest inputs may help in finding the clusters, thus providing useful informationabout the classification labels. The same predictions would not be obtainablefrom a model which induces a function based only on the training cases. Somepeople may call this an example of the closely related semi-supervised learning, since Vapnik's motivation is quite different. The most well-known example of a case-bases learning algorithm is the k-nearest neighbor algorithm, which is related to transductive learning algorithms.[2] Another example of an algorithm in this category is the Transductive Support Vector Machine (TSVM).
A third possible motivation of transduction arises through the needto approximate. If exact inference is computationally prohibitive, one may atleast try to make sure that the approximations are good at the test inputs. Inthis case, the test inputs could come from an arbitrary distribution (notnecessarily related to the distribution of the training inputs), which wouldn'tbe allowed in semi-supervised learning. An example of an algorithm falling inthis category is the Bayesian Committee Machine (BCM).
The mode of inference from particulars to particulars, which Vapnik came to call transduction, was already distinguished from the mode of inference from particulars to generalizations in part III of the Cambridge philosopher and logician W.E. Johnson's 1924 textbook, Logic. In Johnson's work, the former mode was called 'eduction' and the latter was called 'induction'. Bruno de Finetti developed a purely subjective form of Bayesianism in which claims about objective chances could be translated into empirically respectable claims about subjective credences with respect to observables through exchangeability properties. An early statement of this view can be found in his 1937 La Prévision: ses Lois Logiques, ses Sources Subjectives and a mature statement in his 1970 Theory of Probability. Within de Finetti's subjective Bayesian framework, all inductive inference is ultimately inference from particulars to particulars.
The following example problem contrasts some of the unique properties of transduction against induction.
A collection of points is given, such that some of the points are labeled (A, B, or C), but most of the points are unlabeled (?). The goal is to predict appropriate labels for all of the unlabeled points.
The inductive approach to solving this problem is to use the labeled points to train a supervised learning algorithm, and then have it predict labels for all of the unlabeled points. With this problem, however, the supervised learning algorithm will only have five labeled points to use as a basis for building a predictive model. It will certainly struggle to build a model that captures the structure of this data. For example, if a nearest-neighbor algorithm is used, then the points near the middle will be labeled "A" or "C", even though it is apparent that they belong to the same cluster as the point labeled "B".
Transduction has the advantage of being able to consider all of the points, not just the labeled points, while performing the labeling task. In this case, transductive algorithms would label the unlabeled points according to the clusters to which they naturally belong. The points in the middle, therefore, would most likely be labeled "B", because they are packed very close to that cluster.
An advantage of transduction is that it may be able to make better predictions with fewer labeled points, because it uses the natural breaks found in the unlabeled points. One disadvantage of transduction is that it builds no predictive model. If a previously unknown point is added to the set, the entire transductive algorithm would need to be repeated with all of the points in order to predict a label. This can be computationally expensive if the data is made available incrementally in a stream. Further, this might cause the predictions of some of the old points to change (which may be good or bad, depending on the application). A supervised learning algorithm, on the other hand, can label new points instantly, with very little computational cost.
Transduction algorithms can be broadly divided into two categories: those that seek to assign discrete labels to unlabeled points, and those that seek to regress continuous labels for unlabeled points. Algorithms that seek to predict discrete labels tend to be derived by adding partial supervision to a clustering algorithm. Two classes of algorithms can be used: flat clustering and hierarchical clustering. The latter can be further subdivided into two categories: those that cluster by partitioning, and those that cluster by agglomerating. Algorithms that seek to predict continuous labels tend to be derived by adding partial supervision to a manifold learning algorithm.
Partitioning transduction can be thought of as top-down transduction. It is a semi-supervised extension of partition-based clustering. It is typically performed as follows:
Consider the set of all points to be one large partition. While any partition P contains two points with conflicting labels: Partition P into smaller partitions. For each partition P: Assign the same label to all of the points in P.
Of course, any reasonable partitioning technique could be used with this algorithm. Max flow min cut partitioning schemes are very popular for this purpose.
Agglomerative transduction can be thought of as bottom-up transduction. It is a semi-supervised extension of agglomerative clustering. It is typically performed as follows:
Compute the pair-wise distances, D, between all the points. Sort D in ascending order. Consider each point to be a cluster of size 1. For each pair of points in D: If (a is unlabeled) or (b is unlabeled) or (a and b have the same label) Merge the two clusters that contain a and b. Label all points in the merged cluster with the same label.
Manifold-learning-based transduction is still a very young field of research.