Taken the data from quandl API then taking the data having three rows like min max and average. After than categorizing the data using min- max function. Then prediction csn be done using the LSTM

The entire project is divided into 3 modules:1) Socket Programming: The data entered by the users are stored in file and the files are sent toserver where server reads the data and performs various operation as per the requirements.2) AES Encryption: The AES algorithm consists of ten rounds of encryption. First the 128-bit key isexpanded into eleven so-called round keys, each of them 128 bits in size. Each round includes atransformation using the corresponding cipher key to ensure the security of the encryption.3) File Handling: File system here is to store the data that we take from the user. Here, all the datafrom the user is stored in a text file then after going through some processes these files are send toserver via socket and then, at server this file goes through a number of processes to retrieve the data.

1. Design a system to detect traffic congestion.2. Use the system to find alternatives and reduce traffic load.3. Prevent or reduce accidents.Taking the input from Arduino set up in the form of 0 and 1, we will take the signals and then take theinput and store it through SQL in a database.

APF is Authorized program facility, it allows the installation to identity system or user programs that usefunctions to maintain security and integrity, program is to maintain and update APF list throughmainframe (PROGx) member to take input to add and delete the dataset names.

i) Collection of dataset from Quandl API ii) Feature plotting and selectioniii) Normalization of data using min-max and construction of predictive model.iv) Data processing is done by using LSTM and at final stage the prediction and accuracy of dataset.

i) Mechanism to detect the transaction is fraud or not, so Keras sequential model is implemented withRelu and Sigmoid as Activation function. Optimizer is adam.ii) Transactional data has been classified in two categories: 1 for fraud and 0 for correct transaction.