Deep Learning Basics: Exploring Programs & Neural Networks - Essay Sample

Introduction

The first video presents the deep learning basics and its broader scope. This concise video shows vividly several programs on writing tasks that introduce the viewer to use deep learning techniques in areas such as natural language processing, artificial intelligence, and reinforcement learning. It is called deep learning because it creates deep neural networks. It shows the main components which are used in building neural networks and how RapidMiner can enhance one to use Microsoft Cognitive Toolkit, Tensorflow, and the other frameworks which exist in the RapidMiner tool of analysis (Philipp, 2018). I have learned that deep learning can solve complex problems in representation through the introduction of representation that is represented through the use of other representations that are simpler. I have learned the basics of deep learning by working through the learning framework. Additionally, I have learned how to apply the feed-forward, recurrent, and convolution neural nets (Goodfellow et al., 2016). Other things that I have learned are how to embed words, sequence-to-sequence, models that are not supervised, and reinforcement learning.

Is your time best spent reading someone else’s essay? Get a 100% original essay FROM A CERTIFIED WRITER!

Save 25%

Through this video, one can input X and then use it to predict the output. Inform instance, if one is given the stock prices the previous week, he/she can predict the prices of the coming week using deep learning. Finally, if one has a large set of data input and the pairs of output, through deep learning, I can minimize the difference between the expected output and its prediction. Therefore, one can learn the pattern of association between the given outputs and inputs (Philipp, 2018). For one to find the association between the output and input, he/she has to use neural networks. The neural network has the output, the input, and the hidden layers. All three comprises nodes. The numerical representation of data is taken in by the production (pixel specs and images). The output layer's work is to make predictions, and the hidden layers are correlated with the computations. The information is passed between the networks of layers. As the neural networks pass information to the outputs, evaluation of the information on how good the prediction is also made by the network (Philipp, 2018). The evaluation ids did relative to the output through a process called `loss function.'

Lessons from this Video

The video teaches that creativity and imagination in business are important because it can solve complex problems. Since it is rather difficult for humans to define all the rules manually, there is a need to acquire new knowledge that will be important in solving business problems. The traditional way of doing things is steadily being replaced by machines which are making work efficient. The idea presented in this video is so relevant in business because it can be used to predict one's future (Philipp, 2018). Deep learning can be applied in business analytics It is the field that is undergoing a significant transformation due to deep learning. The deep neural can improve the performance predictions when compared with the traditional models. Through this model, business organizations can capitalize on their revenue and profits, identify opportunities, and make necessary adjustments. Deep learning can make a business grow because, through the model, the top management can identify opportunities as well as the threats and make the necessary adjustments for the business's survival (Philipp,2018). Studies also indicate that the companies which use deep learning improve their profits.

The following are the areas where deep learning can be used:

• Analyzing fraud
• Reduction of the costs of inventory
• Analysis of the market basket
• We are analyzing the risks of the business.

An Introduction to Auto Model

The model demonstrated in the second video is different from the existing approaches on the automated machine learning. The model generates the studio processes for Rapid Miner behind the scenes for the data scientists to fine-tune and test the models before they are put in the production stage. The founder of this model is Dr. Ingo Mierswa. In the series of the video demonstrated by Ingo Mierswa, he shows the automated data science from the beginning until the end. I have learned several things from this video (Ingo, 2018). Firstly, it is essential to prepare well because you execute any plan.

Secondly, there are several stages involved before something is produced. Therefore, before a plan is executed, one must be very sure of all the processes involved and ensure that everything is in order before moving to the next stage. For instance, in this automated model, he begins with the preparation of data. Then in the second part, he shows how the machine learning models are built and how they are supposed to be managed (Ingo, 2018). .In the preparation stage, which is the essential part of this model and any other undertaking, he demonstrates how data is loaded, how it is explored, and how the data sets are mashed up. Additionally, this part shows how the data is transformed. All these steps are vital because they will help in the building of good models. The second part of the video is an overview of how the RapidMiner Auto Model works. He then shows how to build the models using the data that was cleaned at the preparation stage. At this stage, it goes through different parameters and values to find the best models. This stage is interconnected with the first stage; therefore, no step can stand on its own. The last part of the video is how to put the model into production, manage the production, and ensure that all the management processes are in place (Ingo, 2018). The three parts show any idea has strict procedures that have to be followed for it to succeed, and the processes are interrelated.

Sentiment Analysis of 700,000 Tweets from Super Bowl 50 - RapidMiner & Aylien

This video shows a technique used by RapidMiner to mine data from the unstructured data. It has a suite used for text mining where things such as transforming cases, tokenizing the words, filtering of the stock words, and stemming are done. The suite helps prepare data into a fashion that is statistically-based for use by a clustering algorithm or may be classified with a linear SVM. They have a partner by the name AYLIEN whose primary work is the addition of natural language processing. The text mining suite can analyze volumes of data that are captured from thousands of tweets. The data is vital in predicting the customers' sentiments and the trends the business brand is likely to take (RapidMiner & Aylien, 2016). Furthermore, the suite can combine all the information with other sources that are structured and create an overall picture of any sentiments or feedback from the customers.

I have learned that if the feedback from the customers is not taken seriously, it can make a business collapse. Therefore, the customers' data analysis is essential in making business decisions that are going to favor the customers. More than 80% of the business data is not structured, and they are contained in thousands of social conversations that take place daily. In understanding the opinions and the needs of the customers, it is essential to unlocking the value hidden in these messages. When a business enterprise understands customer needs, it will be easier for the top management team to make customer-centric decisions. When it comes to data analytics techniques and other data strategies, 90% of this data is not utilized. This happens because it is not easier for human beings to consume vast sets of data, which is presented in a manner that is not structured (RapidMiner & Aylien, 2016). Additionally, data is present at a very high rate making it impossible for the business to analyze and utilize it.

References

Goodfellow, I., Bengio, Y., & Courville, A. (2016). Deep learning. Cambridge, Massachusetts; London, England: MIT Press.

Ingo, M. (2018).An Introduction to Auto Model | RapidMiner.Retrieved fromhttps://www.youtube.com/watch?v=uq36VLMju-4

Philipp, S. (2018).An Introduction to Deep Learning | RapidMiner, Inc. Retrieved from https://www.youtube.com/watch?v=rJCU8ODRwyg&t=3030s

RapidMiner & Aylien. (2016).Sentiment Analysis of 700,000 tweets from Super Bowl 50.Retrieved from https://www.youtube.com/watch?v=CJaqhYL3rY4