Explain binary exponential back-off algorithm.                                   Or  Discuss any one algorithm for allocating a multiple access channel. To ensure that random back-off maintains stability, CSMA/CD uses a technique known as binary exponential back-off. Binary exponential back-off algorithm varies the limit on number of slot times. It works in the following manner- 1) if a stations frame collides for the first time, wait 0 or 1 time slot before trying again. 2) In case of colliding a second time, wait 0, 2 or 3 slot times. 3) If it collides third time, wait anywhere from 0 to 7 slots. 4) After 16 collisions, give up and report an error. Thus, as congestion increase, stations back-off by larger and larger amounts to reduce the probability of collision. This method clearly attempts to reduce excessive waits by keeping the number of possible time slots small. After all if two stations collide, there is a 50% opportunity they will succeed on the next attempt. If several stati

Difference between OSI Model and TCP/IP Model.   Or Difference between Open System Interconnection Model and Transmission Control Protocol/Internet Protocol Model.    1 2  

Overfitting in machine learning  Overfitting refers to a model that models the training data too well. Overfitting happen when a model learn the detail and noise in the training data to the extent that it negatively impacts the performance of the model an new data. This means that the noise or random fluctuations in the training data is picked up and learned as concept by the model. The problem is that these concepts do not apply to new data and negatively impact the model ability to generalize. For example, decision tree are a nonparametric machine learning algorithm that is very flexible and is subject to overfitting training data. This problem can be addressed by pruning a tree after it has learned in order to remove some of the detail it has picked up. Technique to reduce overfitting:- There are several techniques to reduce overfitting in machine learning. 1) Cross Validation 2) L1/L2 regulariszation 3) Feature Selection 4) Hold Out 5) Dropout 1) Cross Validation:- One of the most

Protocol Verification Using Finite State Machine Model:- The finite state machine is a mainly used in several protocol models with this technique, every protocol machine is always in a particular state at every time instant. Variables, including the program counter. From every state, there are zero or more possible transition to other states. When some event happens, transition take place. In case of a protocol machine, a transition might take place if a frame is transmitted, if a frame arrives, if a timer expires, if an interrupt takes place, etc. One particular state is designated as the initial state. This state corresponds to the description of the system when it starts running, or at some convenient starting place shortly thereafter. From the initial state, some, perhaps all, of the other states may be approached by a sequence of transition. this technique is called reachability analysis. This analysis can be helpful in determining whether a protocol is correct. All processes are

 Model and Hyper Parameters  Both Parameters ( model and hyper parameters) are very important. Model Parameter:- It is a configuration variable that is internal to the model and whose value can be estimated from the given data. * They are required by the model when making predictions. * Their values define the skill of the model on your problem. * They are estimated or learned from data. * They are often not set manually by the practitioner. * They are often saved as part of the learned model. Examples:- * The weights in an artificial neural network. * The support vectors in a support vector machine. * The coefficients in a linear regression or logistic regression. Hyper Parameter:- The parameters of the convolution layer which need to be set by the user. Period to the filler learning are called hyper parameters. Hyper Parameters are very important concepts in machine learning especially in the context of neural network since these types of models employ a variety of them. The most imp