METHODOLOGY/DISCUSSION

The proposed methodology and detailed discussion are carried in the present section. A model diagram is presented to highlight how quantum machine learning algorithms can make healthcare smart [75].

The power of quantum computers to manipulate large numbers of highdimensional vectors makes them natural systems for performing vector-based machine learning tasks [71]. Operations that involve taking vector dot products, overlaps, norms, etc., in N-dimensional vector spaces, take O(N) time in the classical machine learning algorithms but the same operations take 0(log N) in the quantum version. These abilities, combined with the quantum linear systems algorithm [25], represent a powerful suite of tools for manipulating large amounts of data. When the size of data increases exponentially and traditional systems get handicapped, quantum computers come as a Messiah to solve this huge data (big data) problem in a respectful time. Currently, the rate of generation of electronic data generated per year is estimated to be on the order of 10¹⁸ bits. This entire data set could be represented by a quantum state using 60 bits, and the clustering analysis could be performed using a few hundred operations. Even if the number of bits to be analyzed were to expand to the entire information content of the universe within the particle horizon 0(Ю⁹⁰ и 2³⁰⁰) bits, in principle, the data representation and analysis would be well within the capacity of a relatively small quantum computer. The generic nature of the quantum speed-ups for dealing with large numbers of high dimensional vectors suggests that a wide variety of machine learning algorithms may be susceptible to exponential speed-up on a quantum computer.

Quantum machines have the ability to search very largely [72] unsorted data sets very quickly. A true quantum computer could encode information in qubits that can be 0 and 1 at the same time. Doing so could reduce the required time to solve a difficult problem that would otherwise take several years of computation to mere seconds. Quantum machine learning methods for analyzing large data sets (big quantum data) supply significant advantages in terms of privacy for the owners of that data in addition to supplying an exponential speedup over classical machine learning algorithms. To store a single number between 0 and 256, a conventional computer uses 8 bit just whereas in a quantum computer 8 qubits can store 256 numbers at once, which dramatically speeds up the computation power. Let us consider all possible combinations of a 2-bit data system with four possible states: 00, 01, 10, and 11. A 2-bit classic computer can at the most simultaneously perform one of these four possible functions. In order to check all of them, the computer would have to repeat each operation separately, whereas a 2-bit quantum computer due to the phenomenon of superposition is able to analyze all of these possibilities simultaneously in one operation. This is due to the fact that 2 qubits contain information about four states whereas 2 bit contains information about one state:

Thus a machine with n qubits can be in a superposition of 2" states at the same time. A 4 qubit computer could analyze 16 parallel states in a single operation; in comparison, a 4-bit classical computer can only analyze one state. To achieve the same solution as the quantum computer classic computer has to repeat this operation 16 times.

• 10 Qubits—Can store 1024 numbers.
• 11 Qubits—Can store 2048 numbers.

In the same way if we are having 100 Qubits, we can be able to store 1, 267,650,600, 228, 229, 401, 496703205, 376 numbers. We can say a quantum computer can tackle the problem on a scale beyond any conventional computer. In this way, we find that quantum computation can be utilized on a vast scale for analyzing the problem of rapidly growing data on the web called big data.

Since big data is mainly characterized by volume, velocity, and variety (3Vs). A major portion of the research is going on to neutralize the issue of variety.

FIGURE 13.4 A framework for using quantum computing for big data in healthcare.

Nowadays, in every sector, we have been able to sense the machines and collect data from them (Streams), known as machine to machine (M2M) data. Data comes in a lot of formats like text, streams, images, sensors, and social media, etc. To draw a useful and meaningful insight from these mountains of data pulps compel to integrate these different formats of data for training the machine, instead of focusing just on one data format. We have proposed a model given in Figure 13.4 in which we have taken the healthcare data as an example. The given framework consists of a cluster of quantum computers connected to a central master quantum computer. The Hadoop/Map-Reduce framework is there to map the data to different nodes in the cluster and reduce the result at the end. In this model firstly data from different sources is inputted to a master node quantum computer, on which Map/Reduce program is been written in Pig or Hive to map the data in the cluster. Since every quantum computer works on the laws of quantum mechanics by working on qubits, which are the lowest levels for data and are similar to all the very different types of data. Thus qubits at electronic level remove the heterogeneity and variety of the data. A /("-means Unsupervised Classification method finally classifies the data among different classes according to its similarity measure. Finally, every group is matched with a predefined threshold value to classify it in Yes or No category indicating the percentage of the presence or absence of that particular target disease [76].