Good one from tmforum-- thought to keep in my blog here, a good cue in data analytic and machine learning from guava's Chief scientist.
"Traditionally, most software companies base their IP on classic computer science algorithms, which pertain to computer architectures, data structures, manipulation and machine usage, etc. This extends to machine learning (ML) where it is often tempting to adopt a “big data” mentality, wherein one starts with whatever data is available and applies various modeling algorithms to see what emerges."
Today, ML is only as good as the data you use to train it. Machine Intelligence (MI) is the next step and it represents the ability of the machine to extrapolate from raw, disparate data to create new valid information that cannot be discovered by applying a machine-learned model.
"CSPs have come to realize the value of analytics. We are now seeing diversity in the use cases, which lead to a definitive ROI. One use case that has risen in popularity over the last few months is network resource optimization. By applying analytics to predict what the network utilization will be at a given time of day, CSPs can determine how best to dynamically deliver high bandwidth consuming applications over their edge networks. This extends the optimizations beyond the reach of CDNs (content delivery networks), for example. For mobile operators, this means that the analytics can pinpoint how subscriber-specific data traffic and radios in RANs (radio access networks) should be managed at any given moment in order to optimize the experiences of all subscribers currently connected to each cell tower."
more insight below .....
"Traditionally, most software companies base their IP on classic computer science algorithms, which pertain to computer architectures, data structures, manipulation and machine usage, etc. This extends to machine learning (ML) where it is often tempting to adopt a “big data” mentality, wherein one starts with whatever data is available and applies various modeling algorithms to see what emerges."
Today, ML is only as good as the data you use to train it. Machine Intelligence (MI) is the next step and it represents the ability of the machine to extrapolate from raw, disparate data to create new valid information that cannot be discovered by applying a machine-learned model.
"CSPs have come to realize the value of analytics. We are now seeing diversity in the use cases, which lead to a definitive ROI. One use case that has risen in popularity over the last few months is network resource optimization. By applying analytics to predict what the network utilization will be at a given time of day, CSPs can determine how best to dynamically deliver high bandwidth consuming applications over their edge networks. This extends the optimizations beyond the reach of CDNs (content delivery networks), for example. For mobile operators, this means that the analytics can pinpoint how subscriber-specific data traffic and radios in RANs (radio access networks) should be managed at any given moment in order to optimize the experiences of all subscribers currently connected to each cell tower."
more insight below .....