In general the answer is "yes" and it is done all the time but it largely depends on your required level of accuracy and pov as a human who generally lives less than 100 years.
If a star's lifetime is measured even in one billion (1,000,000,000 or 10^9) ridiculously short for the vast majority of stars for the purpose of illustrating the vastness in time about which we are discussing, even a plus or minus 1% level of accuracy is an unimaginable number of human lifetimes.
Considering you specified "explode" this rules out those longer lasting stars of less mass. Even Eta Carinae which is calculated to have been formed around 4 x 10^6 years ago and is expected to explode in approximately 10^5 years, has "wiggle room" at even that 1% degree of accuracy (and I doubt anyone claims such high degrees of accuracies) is 1000 years, perhaps some 10-15 human lifetimes.
what is your definition of being able to predict a specific star to go supernova ?
if you mean within the next year, 10, 10,000 or 1 million years, then no
The life cycles of stars are just too long
Predicting that we will see a supernova in our Milky Way in the next 50 - 100 has possibilities
OTOH, predicting which stars would be candidates for becoming a supernova is a little easier
This is done by looking at the type of star. Observing their spectra and seeing where they would be placed
on the main sequence chart
See the Hertzsprung - Russell Diagram
For a star to explode it has to be at least 3.5 solar masses (size of our sun)
Betelgeuse I think is a good example http://nightsky.jpl.nasa.gov/news-display.cfm?News_ID=560
Although Astronomers can't predict exactly when it will happen, they believe Betelgeuse is nearing the end of its lifespan. Betelgeuse is shrinking and even though they don't know why exactly, it's speculated to be a series of partial collapses. These collapses happen when a red giant burns out its hydrogen fuel and switches to other elements like helium and carbon.
By my understanding, only through observed fluctuations such as this can astronomers begin to have an idea of a range of when the individual star might die.
Well, if you could detect silicon fusion occurring in the core of a star, you would know it only has about a day left before it goes boom. Similarly, if oxygen burning was occurring, it would have 6-12 months to live. The trick is being able to discern the nature of fusion occurring in the core of a star. Since the temperature required for either form of burning to occur is known, a way to measure the core temperature of a star would be helpful. Unfortunately, that too is beyond our current capabilities.