I ran a logistic regression with like 10 variables (with R) and some of them have high P-values (>0.05). Should we follow the elimination techniques that we follow in multiple linear regression to remove insignificant variables? Or is the method different in logistic regression?
I'm new to this field so please pardon me if this question sounds silly.
Thank you.
I would like to generate odds-ratios or coefficients for various features in my dataset along with their 95% confidence intervals using a logistic regression model.
Since we cannot generate 95% CI values for odds-ratios or coefficients in sklearn logistic regression models, I started to play with statsmodels.
However, I am not seeing any standard errors for the coefficients in my output using a very large dataset that contains 17 dummy coded categorical features and 1 outcome variable - with modest correlation seen for only a couple of features (Person’s r < 0.45).
My code follows below:
import statsmodels.api as sm
X_atr = sm.add_constant(X_atr) #add constant for intercept
logit_model = sm.Logit(y_atr, X_atr) #Create model instance
result = logit_model.fit(method = "bfgs") #Fit model
print(result.summary()) #print results
Here is a sample of my output. I am getting the coefficients - but without their standard errors or 95% CI values. Can somebody suggest how to fix this issue?
We have two prominent functions (or we can say equations) in logistic regression algorithms:
Logistic regression function.
Logit function.
I would like to know:
Which of these equation(s) is/are used in the logistic regression model building process?
At what stage of model building process which of these equation(s) is/are used?
I know that logit function is used to transform probability values (which range b/w 0 and 1) to real number values (which range b/w -Inf to +Inf). I would like to know the real purpose of logit function in logistic regression modeling process.
Here are few queries which are directly related to the purpose of logit function in Logistic regression modeling:
Has Logit function (i.e. Logit equation LN(P/1-P)) being derived from Logistic Regression equation or its the other way around?
What is the purpose of Logit equation in logistic regression equation? How logit function is used in Logistic regression algorithm? Reason for asking this question will get clear after going through point no. 3 & 4.
Upon building a logistic regression model, we get model coefficients. When we substitute these model coefficients and respective predictor values into the logistic regression equation, we get probability value of being default class (same as the values returned by predict()).
Does this mean that estimated model coefficient values are determined
based on the probability values (computed using logistic regression equation not logit equation) which will be inputed to the likelihood function to determine if it maximizes it or not? If this understanding is correct then, where the logit function is used in the entire process of model building.
Assume that - "Neither logit function is used during model building not during predicting the values". If this is the case then why do we give importance to logit function which is used to map probability values to real number values (ranging between -Inf to +Inf).
Where exactly the logit function is used in the entire logistic regression model buidling process? Is it while estimating the model coefficients?
The model coefficient estimates that we see upon running summary(lr_model) are determined using linear form of logistic regression equation (logit equation) or the actual logistic regression equation?
What is the purpose of Logit function?
The purpose of the Logit function is to convert the real space [0, 1] interval to infinity.
If you check math Logit function, it converts real space from [0,1] interval to infinity [-inf, inf].
Sigmoid and softmax will do exactly the opposite thing. They will convert the [-inf, inf] real space to [0, 1] real space.
This is why in machine learning we may use logit before sigmoid and softmax function, since they match perfectly.
I have a training set with about 300000 examples and about 50-60 features and also it's a multiclass with about 7 classes. I have my logistic regression function that finds out the convergence of the parameters using gradient descent. My gradient descent algorithm, finds the parameters in matrix form as it's faster in matrix form than doing separately and linearly in loops.
Ex :
Matrix(P) <- Matrix(P) - LearningRate( T(Matrix(X)) * ( Matrix(h(X)) -Matrix(Y) ) )
For small training data, it's quite fast and gives correct values with maximum iterations to be around 1000000, but with that much training data, it's extremely slow, that with around 500 iterations it takes 18 minutes, but with that much iterations in gradient descent, the cost is still high and it does not predict the class correctly.
I know, I should implement maybe feature selection, or feature scaling and I can't use the packages provided. Language used is R. How do I go about implementing feature selection or scaling without using any library packages.
According to link, you can use either Z-score normalization or min-max scaling method. Both methods scale the data to [0,1] range. Z-score normalization is calculated as
Min-max scaling method is calculated as:
I've read about neural network a little while ago and I understand how an ANN (especially a multilayer perceptron that learns via backpropagation) can learn to classify an event as true or false.
I think there are two ways :
1) You get one output neuron. It it's value is > 0.5 the events is likely true, if it's value is <=0.5 the event is likely to be false.
2) You get two output neurons, if the value of the first is > than the value of the second the event is likely true and vice versa.
In these case, the ANN tells you if an event is likely true or likely false. It does not tell how likely it is.
Is there a way to convert this value to some odds or to directly get odds out of the ANN. I'd like to get an output like "The event has a 84% probability to be true"
Once a NN has been trained, for eg. using backprogation as mentioned in the question (whereby the backprogation logic has "nudged" the weights in ways that minimize the error function) the weights associated with all individual inputs ("outside" inputs or intra-NN inputs) are fixed. The NN can then be used for classifying purposes.
Whereby the math (and the "options") during the learning phase can get a bit thick, it is relatively simple and straightfoward when operating as a classifier. The main algorithm is to compute an activation value for each neuron, as the sum of the input x weight for that neuron. This value is then fed to an activation function which purpose's is to normalize it and convert it to a boolean (in typical cases, as some networks do not have an all-or-nothing rule for some of their layers). The activation function can be more complex than you indicated, in particular it needn't be linear, but whatever its shape, typically sigmoid, it operate in the same fashion: figuring out where the activation fits on the curve, and if applicable, above or below a threshold. The basic algorithm then processes all neurons at a given layer before proceeding to the next.
With this in mind, the question of using the perceptron's ability to qualify its guess (or indeed guesses - plural) with a percentage value, finds an easy answer: you bet it can, its output(s) is real-valued (if anything in need of normalizing) before we convert it to a discrete value (a boolean or a category ID in the case of several categories), using the activation functions and the threshold/comparison methods described in the question.
So... How and Where do I get "my percentages"?... All depends on the NN implementation, and more importantly, the implementation dictates the type of normalization functions that can be used to bring activation values in the 0-1 range and in a fashion that the sum of all percentages "add up" to 1. In its simplest form, the activation function can be used to normalize the value and the weights of the input to the output layer can be used as factors to ensure the "add up" to 1 question (provided that these weights are indeed so normalized themselves).
Et voilà!
Claritication: (following Mathieu's note)
One doesn't need to change anything in the way the Neural Network itself works; the only thing needed is to somehow "hook into" the logic of output neurons to access the [real-valued] activation value they computed, or, possibly better, to access the real-valued output of the activation function, prior its boolean conversion (which is typically based on a threshold value or on some stochastic function).
In other words, the NN works as previously, neither its training nor recognition logic are altered, the inputs to the NN stay the same, as do the connections between various layers etc. We only get a copy of the real-valued activation of the neurons in the output layer, and we use this to compute a percentage. The actual formula for the percentage calculation depends on the nature of the activation value and its associated function (its scale, its range relative to other neurons' output etc.).
Here are a few simple cases (taken from the question's suggested output rules)
1) If there is a single output neuron: the ratio of the value provided by the activation function relative to the range of that function should do.
2) If there are two (or more output neurons), as with classifiers for example: If all output neurons have the same activation function, the percentage for a given neuron is that of its activation function value divided by the sum of all activation function values. If the activation functions vary, it becomes a case by case situation because the distinct activation functions may be indicative of a purposeful desire to give more weight to some of the neurons, and the percentage should respect this.
What you can do is to use a sigmoid transfer function on the output layer nodes (that accepts data ranges (-inf,inf) and outputs a value in [-1,1]).
Then by using the 1-of-n output encoding (one node for each class), you can map the range [-1,1] to [0,1] and use it as probability for each class value (note that this works naturally for more than just two classes).
The activation value of a single output neuron is a linearly weighted sum, and may be directly interpreted as an approximate probability if the network is trained to give outputs a range from 0 to 1. This would tend to be the case if the transfer function (or output function) in both the preceding stage and providing the final output is in the 0 to 1 range too (typically the sigmoidal logistic function). However, there is no guarantee that it will but repairs are possible. Moreover unless the sigmoids are logistic and the weights are constrained to be positive and sum to 1, it is unlikely. Generally a neural network will train in a more balanced way using the tanh sigmoid and weights and activations that range positive and negative (due to the symmetry of this model). Another factor is the prevalence of the class - if it is 50% then a 0.5 threshold is likely to be effective for logistic and a 0.0 threshold for tanh. The sigmoid is designed to push things towards the centre of the range (on backpropogation) and constrain it from going out of the range (in feedforward). The significance of the performance (with respect to the Bernoulli distribution) can also be interpreted as a probability that the neuron is making real predictions rather than guessing. Ideally the bias of the predictor to positives should match the prevalence of positives in the real world (which may vary at different times and places, e.g. bull vs bear markets, e.g. credit worthiness of people applying for loans vs people who fail to make loan payments) - calibrating to probabilities has the advantage that any desired bias can be set easily.
If you have two neurons for two classes, each can be interpreted independently as above, and the halved difference between them can also be. It is like flipping the negative class neuron and averaging. The differences can also give rise to a probability of significance estimate (using the T-test).
The Brier score and its Murphy decomposition give a more direct estimate of the probability that an average answer is correct, while Informedness gives the probability the classifier is making an informed decision rather than a guess, ROC AUC gives the probability a positive class will be ranked higher than a negative class (by a positive predictor), and Kappa will give a similar number that matches Informedness when prevalence = bias.
What you normally want is both a significance probability for the overall classifier (to ensure that you are playing on a real field, and not in an imaginary framework of guestimates) and a probability estimate for a specific example. There are various ways to calibrate, including doing a regression (linear or nonlinear) versus probability and using its inverse function to remap to a more accurate probability estimate. This can be seen by the Brier score improving, with the calibration component reducing towards 0, but the discrimination component remaining the same, as should ROC AUC and Informedness (Kappa is subject to bias and may worsen).
A simple non-linear way to calibrate to probabilities is to use the ROC curve - as the threshold changes for the output of a single neuron or the difference between two competing neurons, we plot the results true and false positive rates on a ROC curve (the false and true negative rates are naturally the complements, as what isn't really a positive is a negative). Then you scan the ROC curve (polyline) point by point (each time the gradient changes) sample by sample and the proportion of positive samples gives you a probability estimate for positives corresponding to the neural threshold that produced that point. Values between points on the curve can be linearly interpolated between those that are represented in the calibration set - and in fact any bad points in the ROC curve, represented by deconvexities (dents) can be smoothed over by the convex hull - probabilistically interpolating between the endpoints of the hull segment. Flach and Wu propose a technique that actually flips the segment, but this depends on information being used the wrong way round and although it could be used repeatedly for arbitrary improvement on the calibration set, it will be increasingly unlikely to generalize to a test situation.
(I came here looking for papers I'd seen ages ago on these ROC-based approaches - so this is from memory and without these lost references.)
I will be very prudent in interpreting the outputs of a neural networks (in fact any machine learning classifier) as a probability. The machine is trained to discriminate between classes, not to estimate the probability density. In fact, we don't have this information in the data, we have to infer it. For my experience I din't advice anyone to interpret directly the outputs as probabilities.
did you try prof. Hinton's suggestion of training the network with softmax activation function and cross entropy error?
as an example create a three layer network with the following:
linear neurons [ number of features ]
sigmoid neurons [ 3 x number of features ]
linear neurons [ number of classes ]
then train them with cross entropy error softmax transfer with your favourite optimizer stochastic descent/iprop plus/ grad descent. After training the output neurons should be normalized to sum of 1.
Please see http://en.wikipedia.org/wiki/Softmax_activation_function for details. Shark Machine Learning framework does provide Softmax feature through combining two models. And prof. Hinton an excellent online course # http://coursera.com regarding the details.
I can remember I saw an example of Neural network trained with back propagation to approximate the probability of an outcome in the book Introduction to the theory of neural computation (hertz krogh palmer). I think the key to the example was a special learning rule so that you didn't have to convert the output of a unit to probability, but instead you got automatically the probability as output.
If you have the opportunity, try to check that book.
(by the way, "boltzman machines", although less famous, are neural networks designed specifically to learn probability distributions, you may want to check them as well)
When using ANN for 2-class classification and logistic sigmoid activation function is used in the output layer, the output values could be interpreted as probabilities.
So if you choosing between 2 classes, you train using 1-of-C encoding, where 2 ANN outputs will have training values (1,0) and (0,1) for each of classes respectively.
To get probability of first class in percent, just multiply first ANN output to 100. To get probability of other class use the second output.
This could be generalized for multi-class classification using softmax activation function.
You can read more, including proofs of probabilistic interpretation here:
[1] Bishop, Christopher M. Neural networks for pattern recognition. Oxford university press, 1995.