My Path In Science And Engineering Until A Year Ago

Posted on by marin.stoytchev

Blogs are intended for the people who would read them, but in my first post I would like to provide some background before moving on to discussing subjects related to Data Science and Machine Learning.

It is in the nature of scientists and engineers wanting to learn new things and to work on exciting projects and products. This has always been a big factor determining how I feel about my work.

First seven years of my career – predominantly wireless channel characterization work

In the beginning of my carrier, learning new things and working on exciting projects were all there. After my graduate studies on electromagnetic wave propagation in random scattering media, I landed my first job at Lucent Technologies in 1999 as a part of a Physics team at Bell Labs. Our team was conducting measurements for validating the theory on the capacity of multi-channel (MIMO) wireless communications. Everything we were working on was new and exciting. I remember going late to sleep and lying in bed thinking about possible new measurements or about analyzing the measurement data and the results from these analyses.

After three years with Lucent Technologies (last year and a half with Agere Systems, a Lucent spin-off), I moved on and joined the Antenna and Propagation Lab with the Intel Wireless Division in San Diego, CA. There, I continued working on new and exciting things and found myself having the same sleepless nights not being able to wait for the morning to come to continue working on my current project or to start a new one.

During these first six years of my career, my work was mostly research oriented with focus on the wireless channel characterization for multi-antenna systems. There was some work on antenna design and implementation, but this part amounted to probably less than 15 % of my work during that time.

Some of the research work I am most proud of:

A more extensive list of some of my scientific journal articles and conference papers is provided at the end of this post.

Middle stage of my career – entering the field of antenna design, development and implementation

In April 2006, I left Intel and joined an antenna startup where the focus of my work was designing and implementing novel antenna solutions for WiFi devices. There, too, things were new and exciting for me. I felt great satisfaction coming up with new antenna designs and prototypes which we demonstrated successfully to our customers.

In the spring of 2008, I joined an antenna company in Arizona, where my work encompassed the full antenna product cycle – from initial design stage to production. This added new things to my work and I kept learning. I learned many practical things and became better and better over time in finding and implementing the best antenna solution for a particular wireless device.

The tail end – having mastered antenna design, development and implementation

Fast forward seven years and I started noticing that my work was not so exciting anymore. Although the wireless devices we were working on were different to some degree, the antenna solutions were not that different. One could use a set of few antenna designs and configurations which can be implemented successfully in a given device and the rest was simply characterizing the antenna solution, performing few optimization iterations and finalizing the product. The worst part was that everybody else was using mostly the same solutions and, thus, work became a rat race. There was no true innovation. The goal became squeezing 5% more of the antenna solution performance using industry well-known tricks. In many cases, these extra 5% antenna performance boost did not matter as the device performance was determined mostly by how good the radio is and the presence or lack of detrimental noise in the system. Going to sleep excited what the next day would bring and waking up eager to go to work was gone. The excitement was replaced with questioning myself why I am going to work today at all.

So, in the spring of 2015, I started thinking about transitioning into a new field, where there are new and exciting things to learn and work on. Data science had started to make waves and I thought that my experience with statistical analysis and modeling for characterizing the wirelesses propagation channel would be a good match for this field, as well. Moreover, I liked the statistical aspects of my work from those early years and the thought of doing something similar was exciting to me. However, as it has probably happened to a lot of us, life’s routine turned out difficult to overcome and I stayed in the antenna business.

In 2017, I joined another antenna company in San Diego hoping that the new environment would bring back the lost excitement. However, soon it became clear that it is difficult for me to find something new and exciting in the antenna field.

A new beginning – journey into Data Science and Machine Learning

Finally, in January 2019 I decided to leave my job and to focus on transitioning into the fields of Data Science and Machine Learning. Since then, I am even more confident in choosing to pursue a career in these fields. These are among the most innovative and fastest growing branches of Science and Engineering. There is a vast amount of structured and unstructured data businesses can tap into to make better decisions which improve people’s lives. Machines are already doing what is not humanly possible – few examples would be browsing through thousands of images or documents in a matter of minutes and selecting the correct ones; recognizing with higher accuracy than medical experts cases of various diseases such as diabetes, different forms of cancer, and so on. And the potential is limitless. The TED2017 talk by Sebastian Thrun provides some great examples: https://www.ted.com/talks/sebastian_thrun_and_chris_anderson_what_ai_is_and_isn_t#t-6327

In this blog I would like to share my experience along my journey into Data Science and Machine Learning. Hopefully, what I share would be informative and helpful to those who have taken or are thinking of taking a similar path. The first few posts will be posted in a short period of time because over the course of this year I had thought a lot about the subjects I have studied and worked on. After these initial posts, the pace of posting will inevitably slow down and will depend on what I will be working on and whether I believe it to be worthy of sharing or not. Until the next post …

Selected journal articles and conference papers

  • Antenna design for portable computers, Caimi, F. M., Witter, D. C., Ioffe, A. S., Stoytchev, M., chapter in book Emerging Technologies in Wireless LANs: Theory, Design, and Deployment (Bing, B., at all), 2008
  • Beyond 3G: Metamaterials applications to the air interface, Stoytchev, M., Gummalla, A., Achour, M., Poilasne, G., Antennas and Propagation Society International Symposium, June 2007
  • Diversity performance of integrated desktop antennas in a home environment, Stoytchev, M., Witter, D. C., IEEE Antennas and Propagation Society International Symposium, July 2005
  • Simulations of diversity gains of multiple omni and directive antennas in Rician channel with varying K-factor, Stoytchev, M., Witter, D. C., IEEE/ACES International Conference, April 2005
  • Statistics of the Temporal Variations in the Wireless Transmission Channel in Indoor Environments, Stoytchev, M., Safar, H. F., in book Wireless Personal Communications (pp.1-10), January 2002
  • Compact antenna arrays for MIMO applications, Stoytchev, M., Safar, H. F., IEEE Vehicular Technology Conference, Fall 2001
  • Communication in a Disordered World,Simon, S. H., Moustakas, A., Stoytchev, M., Safar, H. F., Physics Today, September 2001
  • Joint spatial and temporal characterization of the wideband wireless communication channel for MIMO applications, Stoytchev, M., Raveche, J. B., Safar, H. F., IEEE Radio and Wireless Conference, February 2001
  • Statistics of the MIMO radio channel in indoor environments, Stoytchev, M., Safar, H., Moustakas, A., Simon, S. H., Antennas and Propagation Society International Symposium, February 2001
  • Statistical signatures of photon localization, Chabanov, A., Stoytchev, M., Genack, A. Z., Nature 404, April 2000
  • Delay-time statistics for diffuse waves, van Tiggelen, B. A., Sebbah, P., Stoytchev, M., Genack, A. Z., Physical Review E 59, July 1999
  • Observations of non-Rayleigh statistics in the approach to photon localization, Stoytchev, M., Genack, A. Z., Optics Letters 24, March 1999
  • Statistics of Wave Dynamics in Random Media, Genack, A. Z., Sebbah, P., Stoytchev, M., van Tiggelen, B. A., Physical Review Letters 82, January 1999
  • Measurement of the Probability Distribution of Total Transmission in Random Waveguides, Stoytchev, M., Genack, A. Z., Physical Review Letters 79, July 1997
  • Microwave transmission through a periodic three-dimensional metal-wire network containing random scatterers, Stoytchev, M., Genack, A. Z., Physical Review B 55, April 1997
  • Measurements of Fluctuations and Correlation of the Total Transmission of Microwave Radiation in an Ensemble of Random Dielectric Samples, Stoytchev, M., Garcia, N., Genack, A. Z., American Physical Society, Annual March Meeting, March 1996

Related Post