Automatic differentiation


In mathematics and computer algebra, automatic differentiation, also called algorithmic differentiation or computational differentiation, is a set of techniques to numerically evaluate the derivative of a function specified by a computer program. AD exploits the fact that every computer program, no matter how complicated, executes a sequence of elementary arithmetic operations and elementary functions. By applying the chain rule repeatedly to these operations, derivatives of arbitrary order can be computed automatically, accurately to working precision, and using at most a small constant factor more arithmetic operations than the original program.
Automatic differentiation is distinct from symbolic differentiation and numerical differentiation.
Symbolic differentiation can lead to inefficient code and faces the difficulty of converting a computer program into a single expression, while numerical differentiation can introduce round-off errors in the discretization process and cancellation. Both classical methods have problems with calculating higher derivatives, where complexity and errors increase. Finally, both classical methods are slow at computing partial derivatives of a function with respect to many inputs, as is needed for gradient-based optimization algorithms. Automatic differentiation solves all of these problems.

The chain rule, forward and reverse accumulation

Fundamental to AD is the decomposition of differentials provided by the chain rule. For the simple composition
the chain rule gives
Usually, two distinct modes of AD are presented, forward accumulation and reverse accumulation. Forward accumulation specifies that one traverses the chain rule from inside to outside, while reverse accumulation has the traversal from outside to inside. More succinctly,
  1. forward accumulation computes the recursive relation: with, and,
  2. reverse accumulation computes the recursive relation: with .
Generally, both forward and reverse accumulation are specific manifestations of applying the operator of program composition, fixing the appropriate one of the two mappings.

Forward accumulation

In forward accumulation AD, one first fixes the independent variable with respect to which differentiation is performed and computes the derivative of each sub-expression recursively. In a pen-and-paper calculation, this involves repeatedly substituting the derivative of the inner functions in the chain rule:
This can be generalized to multiple variables as a matrix product of Jacobians.
Compared to reverse accumulation, forward accumulation is natural and easy to implement as the flow of derivative information coincides with the order of evaluation. Each variable is augmented with its derivative ,
as denoted by the dot. The derivatives are then computed in sync with the evaluation steps and combined with other derivatives via the chain rule.
As an example, consider the function:
For clarity, the individual sub-expressions have been labeled with the variables.
The choice of the independent variable to which differentiation is performed affects the seed values and. Given interest in the derivative of this function with respect to, the seed values should be set to:
With the seed values set, the values propagate using the chain rule as shown. Figure 2 shows a pictorial depiction of this process as a computational graph.
To compute the gradient of this example function, which requires the derivatives of with respect to not only but also, an additional sweep is performed over the computational graph using the seed values.
The computational complexity of one sweep of forward accumulation is proportional to the complexity of the original code.
Forward accumulation is more efficient than reverse accumulation for functions with as only sweeps are necessary, compared to sweeps for reverse accumulation.

Reverse accumulation

In reverse accumulation AD, the dependent variable to be differentiated is fixed and the derivative is computed with respect to each sub-expression recursively. In a pen-and-paper calculation, the derivative of the outer functions is repeatedly substituted in the chain rule:
In reverse accumulation, the quantity of interest is the adjoint, denoted with a bar ; it is a derivative of a chosen dependent variable with respect to a subexpression :
Reverse accumulation traverses the chain rule from outside to inside, or in the case of the computational graph in Figure 3, from top to bottom. The example function is scalar-valued, and thus there is only one seed for the derivative computation, and only one sweep of the computational graph is needed to calculate the gradient. This is only half the work when compared to forward accumulation, but reverse accumulation requires the storage of the intermediate variables as well as the instructions that produced them in a data structure known as a Wengert list, which may consume significant memory if the computational graph is large. This can be mitigated to some extent by storing only a subset of the intermediate variables and then reconstructing the necessary work variables by repeating the evaluations, a technique known as rematerialization. Checkpointing is also used to save intermediary states.
The operations to compute the derivative using reverse accumulation are shown in the table below :
The data flow graph of a computation can be manipulated to calculate the gradient of its original calculation. This is done by adding an adjoint node for each primal node, connected by adjoint edges which parallel the primal edges but flow in the opposite direction. The nodes in the adjoint graph represent multiplication by the derivatives of the functions calculated by the nodes in the primal. For instance, addition in the primal causes fanout in the adjoint; fanout in the primal causes addition in the adjoint; a unary function in the primal causes in the adjoint; etc.
Reverse accumulation is more efficient than forward accumulation for functions with as only sweeps are necessary, compared to sweeps for forward accumulation.
Reverse mode AD was first published in 1976 by Seppo Linnainmaa.
Backpropagation of errors in multilayer perceptrons, a technique used in machine learning, is a special case of reverse mode AD.

Beyond forward and reverse accumulation

Forward and reverse accumulation are just two ways of traversing the chain rule. The problem of computing a full Jacobian of with a minimum number of arithmetic operations is known as the optimal Jacobian accumulation problem, which is NP-complete. Central to this proof is the idea that algebraic dependencies may exist between the local partials that label the edges of the graph. In particular, two or more edge labels may be recognized as equal. The complexity of the problem is still open if it is assumed that all edge labels are unique and algebraically independent.

Automatic differentiation using dual numbers

Forward mode automatic differentiation is accomplished by augmenting the algebra of real numbers and obtaining a new arithmetic. An additional component is added to every number to represent the derivative of a function at the number, and all arithmetic operators are extended for the augmented algebra. The augmented algebra is the algebra of dual numbers. This approach was generalized by the theory of operational calculus on programming spaces, through tensor algebra of the dual space.
Replace every number with the number, where is a real number, but is an abstract number with the property . Using only this, regular arithmetic gives
and likewise for subtraction and division.
Now, polynomials can be calculated in this augmented arithmetic. If, then
where denotes the derivative of with respect to its first argument, and, called a seed, can be chosen arbitrarily.
The new arithmetic consists of ordered pairs, elements written, with ordinary arithmetics on the first component, and first order differentiation arithmetic on the second component, as described above. Extending the above results on polynomials to analytic functions gives a list of the basic arithmetic and some standard functions for the new arithmetic:
and in general for the primitive function,
where and are the derivatives of with respect to its first and second arguments, respectively.
When a binary basic arithmetic operation is applied to mixed arguments—the pair and the real number —the real number is first lifted to. The derivative of a function at the point is now found by calculating using the above arithmetic, which gives as the result.

Vector arguments and functions

Multivariate functions can be handled with the same efficiency and mechanisms as univariate functions by adopting a directional derivative operator. That is, if it is sufficient to compute, the directional derivative of at in the direction, this may be calculated as using the same arithmetic as above. If all the elements of are desired, then function evaluations are required. Note that in many optimization applications, the directional derivative is indeed sufficient.

High order and many variables

The above arithmetic can be generalized to calculate second order and higher derivatives of multivariate functions. However, the arithmetic rules quickly grow complicated: complexity is quadratic in the highest derivative degree. Instead, truncated Taylor polynomial algebra can be used. The resulting arithmetic, defined on generalized dual numbers, allows efficient computation using functions as if they were a data type. Once the Taylor polynomial of a function is known, the derivatives are easily extracted. A rigorous, general formulation is achieved through the tensor series expansion using operational calculus on programming spaces.

Operational calculus on programming spaces

on programming spaces provides differentiable programming with formal semantics through an algebra of higher-order constructs. It can thus be used to express the concepts underlying automatic differentiation.

Differentiable programming space

A differentiable programming space is any subspace of such that
where is the tensor algebra of the dual space. When all elements of are analytic, we call an analytic programming space.
Definition. Let be a differentiable programming space. The space spanned by over, where is called a differentiable programming space of order.
This means that calculation of derivatives of the map, can be calculated with only one mapping. The operator as a direct sum of operators
The image is a multitensor of order, which is a direct sum of the maps value and all derivatives of order, all evaluated at the point
The operator satisfies the recursive relation.
that can be used to recursively construct programming spaces of arbitrary order. Only explicit knowledge of is required for the construction of from, which is evident from the above theorem.

Virtual tensor machine

An abstract virtual machine is capable of constructing and implementing the theory. Such a machine provides a framework for analytic study of algorithmic procedures through algebraic means.
This claim allows a simple definition of such a machine.
Definition. The tuple is a virtual tensor machine, where
Expansion into a series offers valuable insights into programs through methods of analysis.
There exists a space spanned by the set over a field. Thus, the expression
is well defined. The operator is a mapping between function spaces
It also defines a map
by taking the image of the map at a certain point.
A map can be constructed from the space of programs to the space of polynomials. The space of multivariate polynomials is isomorphic to symmetric algebra, which is in turn a quotient of tensor algebra. To any element of the corresponding element of can be attached, namely a polynomial map. Thus, the completion of the symmetric algebra is the formal power series, which is in turn isomorphic to a quotient of tensor series algebra, arriving at
For any element, the expression is a map, mapping a program to a formal power series. The correspondence between multi-tensors in and polynomial maps can be given by multiple contractions for all possible indices.
Evaluated at, the operator is a generalization of the shift operator widely used in physics. For a specific it is here on denoted by
When the choice of is arbitrary, it is omitted from expressions for brevity. Others adopted similar approaches.

Operator of program composition

Theory offers a generalization of both forward and reverse mode of automatic differentiation to arbitrary order, under a single invariant operator in the theory. This condenses complex notions to simple expressions allowing meaningful manipulations before being applied to a particular programming space.
Proof can be found in.
Both forward and reverse mode are obtainable using this operator, by fixing the appropriate one of the two maps. This generalizes both concepts under a single operator in the theory. For example, by considering projections of the operator onto the space spanned by, and fixing the second map, we retrieve the basic first order forward mode of automatic differentiation, or reverse mode, by fixing.
Thus the operator alleviates the need for explicit implementation of the higher order chain rule, as it is encoded in the structure of the operator itself, which can be efficiently implemented by manipulating its generating map.

Order reduction for nested applications

It is useful to be able to use the -th derivative of a program as part of a different differentiable program. As such, we must be able to treat the derivative itself as a differentiable program, while only coding the original program.
By the above Theorem, -differentiable -th derivatives of a program can be extracted by
Thus, we gained the ability of writing a differentiable program acting on derivatives of another program, stressed as crucial by other authors.

Implementation

Forward-mode AD is implemented by a nonstandard interpretation of the program in which real numbers are replaced by dual numbers, constants are lifted to dual numbers with a zero epsilon coefficient, and the numeric primitives are lifted to operate on dual numbers. This nonstandard interpretation is generally implemented using one of two strategies: source code transformation or operator overloading.

Operator overloading (OO)

is a possibility for source code written in a language supporting it. Objects for real numbers and elementary mathematical operations must be overloaded to cater for the augmented arithmetic depicted above. This requires no change in the form or sequence of operations in the original source code for the function to be differentiated, but often requires changes in basic data types for numbers and vectors to support overloading and often also involves the insertion of special flagging operations.
Operator overloading for forward accumulation is easy to implement, and also possible for reverse accumulation. However, current compilers lag behind in optimizing the code when compared to forward accumulation.
Operator overloading, for both forward and reverse accumulation, can be well-suited to applications where the objects are vectors of real numbers rather than scalars. This is because the tape then comprises vector operations; this can facilitate computationally efficient implementations where each vector operation performs many scalar operations. Vector adjoint algorithmic differentiation techniques may be used, for example, to differentiate values calculated by Monte-Carlo simulation.
Examples of operator-overloading implementations of automatic differentiation in C++ are the Adept and Stan libraries.