I am studying informed search algorithms, and for New Bidirectional A* Search, I know that the space complexity is O(b^d), where d is the depth of the shallowest goal node and b is the branch factor. I have tried to find out what its time complexity is, but I haven't been able to find any exact information about it on online resources. Is the exact time complexity of NBA* Search unknown and what is the difference between the original Bidirectional A*? Any insights are appreciated.
If you have specific models of your problem (eg uniformly growing graph in both directions with unit edge costs and the number of states growing exponentially) then most bidirectional search algorithms require O(b^(d/2)) node expansions and require O(b^(d/2)) time. But, this simple model doesn't actual model most real-world problems.
Given this, I would not recommend putting significant effort into studying New Bidirectional A*.
The state of the art in bidirectional search has changed massively in the last few years. The current algorithm with the best theoretical guarantees is NBS - Near-Optimal Bidirectional Search. The algorithm finds optimal paths and is near-optimal in node-expansions. That is, NBS is guaranteed to do no more than 2x more necessary expansions than the best possible algorithm (given reasonable theoretical assumptions such as using the same heuristic). All other algorithms (including A*) can do arbitrarily worse than NBS.
Other algorithm variants of NBS, such as DVCBS, have been proposed which follow the same basic structure, do not have the same guarantees, but perform well in practice.
Related
In my Intro to AI class, we have been studying:
Uniformed Search (i.e. Depth-First Search)
Informed Search (i.e. A* Search)
Constraint Satisfaction pRoblems (i.e. Hill Climbing)
Adversarial Search (i.e. Minimax)
In general lines, why would we use, for example, Depth-First Search instead of using more complex algorithms such as A* Search? In other words, why choosing simple and limited algorithms when we can choose complex ones?
The main reason is efficiency. Some algorithms take much more time/memory than others.
Some algorithms won't work will in certain situations. For example, if there are local maxima, Hill Climbing won't work very well.
If you expect most paths to lead to destination, you can use Depth First, which could be much faster than A*.
In Artificial Intelligence - A Modern Approach 3rd Edition, I came across an interesting quote stating:
"As yet there is no good understanding of how to combine the two kinds of algorithms [Goal directed reasoning / planning and heuristic search] into a robust and efficient system" (Russel pg 189)
Why is this so? Why is it hard to combine goal oriented planning with a heuristic search? Wouldn't reinforcement learning solve this?
The term “Goal directed reasoning” was used in the 1980s for a backtracking search technique. Sometimes it was called backward reasoning or top-down search, which means all the same. It describes the working of the algorithm in traversing the state space. Or to be more specific: it describes the order in which the states in the graph are visited. In newer literature this aspect of a planner is no longer explained in detail, because a graph search algorithm is no big thing. It means simply to put the nodes on a stack and traverse them.
In contrast, the term “heuristic search” means to replace a brute-force solver with a knowledge based approach. Heuristic search is equal to not traverse a graph, but find a domain-specific strategy which leaves out most part of the graph. And indeed, it is hard to combine backtracking with heuristics, this approach would be called grounding. If a grounded problem is available, it is possible to use a backtracking solver on a knowledge-based problem. This is the strategy utilized in modern PDDL planners which are first describe the domain in a symbolic PDDL notation (which is knowledge based) and using then a fast solver to search in the state space.
What is difference between SVM and Neural Network?
Is it true that linear svm is same NN, and for non-linear separable problems, NN uses adding hidden layers and SVM uses changing space dimensions?
There are two parts to this question. The first part is "what is the form of function learned by these methods?" For NN and SVM this is typically the same. For example, a single hidden layer neural network uses exactly the same form of model as an SVM. That is:
Given an input vector x, the output is:
output(x) = sum_over_all_i weight_i * nonlinear_function_i(x)
Generally the nonlinear functions will also have some parameters. So these methods need to learn how many nonlinear functions should be used, what their parameters are, and what the value of all the weight_i weights should be.
Therefore, the difference between a SVM and a NN is in how they decide what these parameters should be set to. Usually when someone says they are using a neural network they mean they are trying to find the parameters which minimize the mean squared prediction error with respect to a set of training examples. They will also almost always be using the stochastic gradient descent optimization algorithm to do this. SVM's on the other hand try to minimize both training error and some measure of "hypothesis complexity". So they will find a set of parameters that fits the data but also is "simple" in some sense. You can think of it like Occam's razor for machine learning. The most common optimization algorithm used with SVMs is sequential minimal optimization.
Another big difference between the two methods is that stochastic gradient descent isn't guaranteed to find the optimal set of parameters when used the way NN implementations employ it. However, any decent SVM implementation is going to find the optimal set of parameters. People like to say that neural networks get stuck in a local minima while SVMs don't.
NNs are heuristic, while SVMs are theoretically founded. A SVM is guaranteed to converge towards the best solution in the PAC (probably approximately correct) sense. For example, for two linearly separable classes SVM will draw the separating hyperplane directly halfway between the nearest points of the two classes (these become support vectors). A neural network would draw any line which separates the samples, which is correct for the training set, but might not have the best generalization properties.
So no, even for linearly separable problems NNs and SVMs are not same.
In case of linearly non-separable classes, both SVMs and NNs apply non-linear projection into higher-dimensional space. In the case of NNs this is achieved by introducing additional neurons in the hidden layer(s). For SVMs, a kernel function is used to the same effect. A neat property of the kernel function is that the computational complexity doesn't rise with the number of dimensions, while for NNs it obviously rises with the number of neurons.
Running a simple out-of-the-box comparison between support vector machines and neural networks (WITHOUT any parameter-selection) on several popular regression and classification datasets demonstrates the practical differences: an SVM becomes a very slow predictor if many support vectors are being created while a neural network's prediction speed is much higher and model-size much smaller. On the other hand, the training time is much shorter for SVMs. Concerning the accuracy/loss - despite the aforementioned theoretical drawbacks of neural networks - both methods are on par - especially for regression problems, neural networks often outperform support vector machines. Depending on your specific problem, this might help to choose the right model.
Both Support Vector Machines (SVMs) and Artificial Neural Networks (ANNs) are supervised machine learning classifiers. An ANN is a parametric classifier that uses hyper-parameters tuning during the training phase. An SVM is a non-parametric classifier that finds a linear vector (if a linear kernel is used) to separate classes. Actually, in terms of the model performance, SVMs are sometimes equivalent to a shallow neural network architecture. Generally, an ANN will outperform an SVM when there is a large number of training instances, however, neither outperforms the other over the full range of problems.
We can summarize the advantages of the ANN over the SVM as follows:
ANNs can handle multi-class problems by producing probabilities for each class. In contrast, SVMs handle these problems using independent one-versus-all classifiers where each produces a single binary output. For example, a single ANN can be trained to solve the hand-written digits problem while 10 SVMs (one for each digit) are required.
Another advantage of ANNs, from the perspective of model size, is that the model is fixed in terms of its inputs nodes, hidden layers, and output nodes; in an SVM, however, the number of support vector lines could reach the number of instances in the worst case.
The SVM does not perform well when the number of features is greater than the number of samples. More work in feature engineering is required for an SVM than that needed for a multi-layer Neural Network.
On the other hand, SVMs are better than ANNs in certain respects:
In comparison to SVMs, ANNs are more prone to becoming trapped in local minima, meaning that they sometime miss the global picture.
While most machine learning algorithms can overfit if they don’t have enough training samples, ANNs can also overfit if training goes on for too long - a problem that SVMs do not have.
SVM models are easier to understand. There are different kernels that provide a different level of flexibilities beyond the classical linear kernel, such as the Radial Basis Function kernel (RBF). Unlike the linear kernel, the RBF can handle the case when the relation between class labels and attributes is nonlinear.
SVMs and NNs have the same building block as perceptrons, but SVMs also uses a kernel trick to raise dimension from say 2 to 3d by translation such as Y = (x1,2,..^2, y1,2...^2) which can separate linearly inseparable plains using a straight line. Want a demo like it and ask me :)
Actually, they are exactly equivalent to each other. The only difference is in their standard implementations with selections of activation function and regularization etc, which obviously differ from each other. Also, I have yet not seen a dual formulation for neural networks, but SVMs are moving toward the primal anyway.
Practically, most of your assumption are often quite true. I'll elaborate: for linear separable classes Linear SVM works quite good and and it's much faster to train. For non linear classes there is the kernel trick, which is sending your data to a higher dimension space. This trick however has two disadvantages compared to NN. First - your have to search for the right parameters , because the classifier will only work if in the higher dimension the two sets will be linearly separable. Now - testing parameters is often done by grid search which is CPU-time consuming. The other problem is that this whole technique is not as general as NN (for example, for NLP if often results in poor classifier).
after studying decision tree for a while, I noticed there is a small technique called boosting. I see in normal cases, it will improve the accuracy of the decision tree.
So I am just wondering, why don't we just simply incorporate this boosting into every decision tree we built? Since currently we leave boosting out as a separate technique, so I ponder: are there any disadvantages of using boosting than just using a single decision tree?
Thanks for helping me out here!
Boosting is a technique that can go on top any learning algorithm. It is the most effective when the original classifier you built performs just barely above random. If your decision tree is pretty good already, boosting may not make much difference, but have performance penalty -- if you run boosting for 100 iterations you'll have to train and store 100 decision trees.
Usually people do boost with decision stumps (decision trees with just one node) and get results as good as boosting with full decision trees.
I've done some experiments with boosting and found it to be fairly robust, better than single tree classifier, but also slower (I used to 10 iterations), and not as good as some of the simpler learners (to be fair, it was an extremely noisy dataset)
there are several disadvatages for boosting:
1-hard to implement
2-they need extensive training with training sets more than a decision tree does
3- the worst thing is that all boosting algorithms require a Threshold value
which is in most cases not easy to figure out because it requires extensive trial and error tests knowing that the whole performance of the boosting algorithm depends on this threshold
What are the relevant differences, in terms of performance and use cases, between simulated annealing (with bean search) and genetic algorithms?
I know that SA can be thought as GA where the population size is only one, but I don't know the key difference between the two.
Also, I am trying to think of a situation where SA will outperform GA or GA will outperform SA. Just one simple example which will help me understand will be enough.
Well strictly speaking, these two things--simulated annealing (SA) and genetic algorithms are neither algorithms nor is their purpose 'data mining'.
Both are meta-heuristics--a couple of levels above 'algorithm' on the abstraction scale. In other words, both terms refer to high-level metaphors--one borrowed from metallurgy and the other from evolutionary biology. In the meta-heuristic taxonomy, SA is a single-state method and GA is a population method (in a sub-class along with PSO, ACO, et al, usually referred to as biologically-inspired meta-heuristics).
These two meta-heuristics are used to solve optimization problems, particularly (though not exclusively) in combinatorial optimization (aka constraint-satisfaction programming). Combinatorial optimization refers to optimization by selecting from among a set of discrete items--in other words, there is no continuous function to minimize. The knapsack problem, traveling salesman problem, cutting stock problem--are all combinatorial optimization problems.
The connection to data mining is that the core of many (most?) supervised Machine Learning (ML) algorithms is the solution of an optimization problem--(Multi-Layer Perceptron and Support Vector Machines for instance).
Any solution technique to solve cap problems, regardless of the algorithm, will consist essentially of these steps (which are typically coded as a single block within a recursive loop):
encode the domain-specific details
in a cost function (it's the
step-wise minimization of the value
returned from this function that
constitutes a 'solution' to the c/o
problem);
evaluate the cost function passing
in an initial 'guess' (to begin
iteration);
based on the value returned from the
cost function, generate a subsequent
candidate solution (or more than
one, depending on the
meta-heuristic) to the cost
function;
evaluate each candidate solution by
passing it in an argument set, to
the cost function;
repeat steps (iii) and (iv) until
either some convergence criterion is
satisfied or a maximum number of
iterations is reached.
Meta-heuristics are directed to step (iii) above; hence, SA and GA differ in how they generate candidate solutions for evaluation by the cost function. In other words, that's the place to look to understand how these two meta-heuristics differ.
Informally, the essence of an algorithm directed to solution of combinatorial optimization is how it handles a candidate solution whose value returned from the cost function is worse than the current best candidate solution (the one that returns the lowest value from the cost function). The simplest way for an optimization algorithm to handle such a candidate solution is to reject it outright--that's what the hill climbing algorithm does. But by doing this, simple hill climbing will always miss a better solution separated from the current solution by a hill. Put another way, a sophisticated optimization algorithm has to include a technique for (temporarily) accepting a candidate solution worse than (i.e., uphill from) the current best solution because an even better solution than the current one might lie along a path through that worse solution.
So how do SA and GA generate candidate solutions?
The essence of SA is usually expressed in terms of the probability that a higher-cost candidate solution will be accepted (the entire expression inside the double parenthesis is an exponent:
p = e((-highCost - lowCost)/temperature)
Or in python:
p = pow(math.e, (-hiCost - loCost) / T)
The 'temperature' term is a variable whose value decays during progress of the optimization--and therefore, the probability that SA will accept a worse solution decreases as iteration number increases.
Put another way, when the algorithm begins iterating, T is very large, which as you can see, causes the algorithm to move to every newly created candidate solution, whether better or worse than the current best solution--i.e., it is doing a random walk in the solution space. As iteration number increases (i.e., as the temperature cools) the algorithm's search of the solution space becomes less permissive, until at T = 0, the behavior is identical to a simple hill-climbing algorithm (i.e., only solutions better than the current best solution are accepted).
Genetic Algorithms are very different. For one thing--and this is a big thing--it generates not a single candidate solution but an entire 'population of them'. It works like this: GA calls the cost function on each member (candidate solution) of the population. It then ranks them, from best to worse, ordered by the value returned from the cost function ('best' has the lowest value). From these ranked values (and their corresponding candidate solutions) the next population is created. New members of the population are created in essentially one of three ways. The first is usually referred to as 'elitism' and in practice usually refers to just taking the highest ranked candidate solutions and passing them straight through--unmodified--to the next generation. The other two ways that new members of the population are usually referred to as 'mutation' and 'crossover'. Mutation usually involves a change in one element in a candidate solution vector from the current population to create a solution vector in the new population, e.g., [4, 5, 1, 0, 2] => [4, 5, 2, 0, 2]. The result of the crossover operation is like what would happen if vectors could have sex--i.e., a new child vector whose elements are comprised of some from each of two parents.
So those are the algorithmic differences between GA and SA. What about the differences in performance?
In practice: (my observations are limited to combinatorial optimization problems) GA nearly always beats SA (returns a lower 'best' return value from the cost function--ie, a value close to the solution space's global minimum), but at a higher computation cost. As far as i am aware, the textbooks and technical publications recite the same conclusion on resolution.
but here's the thing: GA is inherently parallelizable; what's more, it's trivial to do so because the individual "search agents" comprising each population do not need to exchange messages--ie, they work independently of each other. Obviously that means GA computation can be distributed, which means in practice, you can get much better results (closer to the global minimum) and better performance (execution speed).
In what circumstances might SA outperform GA? The general scenario i think would be those optimization problems having a small solution space so that the result from SA and GA are practically the same, yet the execution context (e.g., hundreds of similar problems run in batch mode) favors the faster algorithm (which should always be SA).
It is really difficult to compare the two since they were inspired from different domains..
A Genetic Algorithm maintains a population of possible solutions, and at each step, selects pairs of possible solution, combines them (crossover), and applies some random changes (mutation). The algorithm is based the idea of "survival of the fittest" where the selection process is done according to a fitness criteria (usually in optimization problems it is simply the value of the objective function evaluated using the current solution). The crossover is done in hope that two good solutions, when combined, might give even better solution.
On the other hand, Simulated Annealing only tracks one solution in the space of possible solutions, and at each iteration considers whether to move to a neighboring solution or stay in the current one according to some probabilities (which decays over time). This is different from a heuristic search (say greedy search) in that it doesn't suffer from the problems of local optimum since it can get unstuck from cases where all neighboring solutions are worst the current one.
I'm far from an expert on these algorithms, but I'll try and help out.
I think the biggest difference between the two is the idea of crossover in GA and so any example of a learning task that is better suited to GA than SA is going to hinge on what crossover means in that situation and how it is implemented.
The idea of crossover is that you can meaningfully combine two solutions to produce a better one. I think this only makes sense if the solutions to a problem are structured in some way. I could imagine, for example, in multi-class classification taking two (or many) classifiers that are good at classifying a particular class and combining them by voting to make a much better classifier. Another example might be Genetic Programming, where the solution can be expressed as a tree, but I find it hard to come up with a good example where you could combine two programs to create a better one.
I think it's difficult to come up with a compelling case for one over the other because they really are quite similar algorithms, perhaps having been developed from very different starting points.